Possible role of streptococcal infection in flares of juvenile idiopathic arthritis



Juvenile idiopathic arthritis (JIA) is defined as onset of disease at <16 years of age, with arthritis persisting in ≥1 joint for at least 6 weeks and with active exclusion of other well-defined diseases such as systemic lupus erythematosus, etc. JIA is a heterogeneous group of arthritides with diverse presentation, course, and outcome (1).

The role of infection, particularly streptococcal infection, in the pathogenesis of autoimmune diseases has been studied in the past (2). There is some evidence that streptococcal infection may be a trigger or maintenance factor in the pathogenesis of juvenile dermatomyositis (DM) (3) and psoriasis (4). In this report, we describe a group of patients with JIA in whom streptococcal infection possibly caused a flare or worsening of their chronic disease.

Patients and Methods

In a pediatric rheumatology service of a local hospital, we observed, during the last 10 years, 9 patients with various types of JIA who experienced repeated disease exacerbation during streptococcal infection (the study group). Streptococcal infection was proven with positive throat culture and/or elevated antistreptolysin O (ASO) titer in all patients at least twice before inclusion. ASO is checked routinely at onset (if the child is seen close to the beginning of the arthritis) and at the time of disease flare or worsening, and throat culture is obtained following either clinical suspicion or high antistreptolysin titer. In the case of positive culture we routinely advise repeated culture after treatment to prove eradication and exclude carrier state. ASO was considered elevated if >500 IU, which is above the upper limit of normal for children that has been found in various studies (5, 6).

Remission was defined as no active joints, no pain according to the patient, and normal erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level. Flare was defined as active joint disease with or without elevated ESR and CRP level after the patient has been in remission for at least 3 months. Worsening of disease was defined according to the criteria proposed by Brunner et al (7). The clinical data of these patients are summarized in Table 1. Patient 1 is described in detail below.

Table 1. Clinical data of patients with JIA and streptococcal infection*
PatientAge at onset, yearsSexType of JIAUveitisANANo. of exacerbations with proven streptococcal infectionPenicillin prophylaxisYears disease free without treatment
  • *

    JIA = juvenile idiopathic arthritis; ANA = antinuclear antibody; oligo = oligoarticular; poly = polyarticular; ext. oligo = extended oligoarticular.

86FExt. oligoNo+4Yes2

The files of all the patients diagnosed with JIA during the same period were reviewed. Thirty-two additional patients (the control group) were studied and their clinical data were summarized and compared with the data of the study group concerning age at onset, type of JIA, presence of positive antinuclear antibody (ANA), uveitis, and time in remission without treatment. JIA was diagnosed according to the International League of Associations for Rheumatology (ILAR) criteria (8), namely, persistence of arthritis for at least 6 weeks in a child <16 years of age, with active exclusion of other known conditions.

HLA–B27 was checked in 2 teenage male patients, one in the study group and another in the control group; both were found to be negative. No other HLA studies were performed.

Data are reported as the mean ± SD. The statistical analysis was performed using SPSS software, version 12 for Windows (SPSS, Chicago, IL). Differences between the study group and controls were tested by 2-sample t-tests for normally distributed data and Wilcoxon's ranked sum tests for data not normally distributed.

Case report of patient 1.

Patient 1 was a young girl first seen at the age of 6, with arthritis of the right knee of a few months' duration, mild uveitis, and positive ANA. After failure of treatment with nonsteroidal antiinflammatory drugs (NSAIDs) for 3 months, her knee was injected with triamcinolone acetonide, resulting in complete recovery. The patient was lost to followup and returned 2 years later at the age of 9 with severe arthritis of both knees and a history of sore throat several weeks earlier. At this time she had high ASO titer (1,600 IU) and positive throat culture for group A beta hemolytic streptococcus (GAS). The patient was treated with antibiotics and an NSAID and was started on penicillin prophylaxis; within a few months remission of the arthritis was achieved and the NSAID was stopped.

Two years later, at the age of 11 the patient presented with bilateral knee arthritis with large effusions of both knees, positive throat culture for GAS, and extremely elevated ASO titer (4,800 IU). This time there was no history of sore throat. The mother reported that 2 months prior to this exacerbation they stopped the penicillin prophylaxis. The patient again was treated with antibiotics followed by penicillin prophylaxis every 3 weeks and intraarticular injection of triamcinolone acetonide to both knees, resulting in prompt resolution of her arthritis. During the last 5 years the patient has adhered to the prophylactic treatment with good results. Three years ago she experienced a mild flare of arthritis in 1 knee associated with proven streptococcal infection, which occurred during a trial to reduce the frequency of the penicillin injections to every 4 weeks. During the last 3 years the patient has been disease free concerning her arthritis and uveitis and receives penicillin prophylaxis every 3 weeks with no other medication.


A total of 41 patients with JIA were diagnosed during the defined period. An association between streptococcal infection and exacerbation of the joint disease was observed in 9 patients, and as a result these patients received penicillin prophylaxis (the study group) (Table 1). Two patients received parenteral benzathine penicillin every 3–4 weeks and the other 7 patients received oral penicillin once daily. The mean number of exacerbations observed with proven GAS infection was 3 (range 2–5). The clinical and laboratory features of the flares are summarized in Table 2. Evidence of streptococcal infection at onset or worsening was investigated in the additional 32 patients that served as the control group. Data on ASO titer were found at onset in 24 of 32 patients; the remaining 8 patients were first seen by us months or years after onset, so there was no point in looking for evidence of streptococcal infection. Of the 24 patients checked for ASO titer, 3 had elevated titers (>500 IU): 727 IU, 700 IU, and 898 IU; none had positive throat culture. There were 25 episodes of flare or worsening in the 32 patients in the control group; evidence of recent streptococcal infection was found in 5 of these patients. Of these 5 patients, 2 had positive throat culture that became negative after antibiotic treatment, 2 had positive throat culture and elevated ASO titer (726 and 1,200 IU), and the fifth had only elevated ASO titer (898 IU). None of the patients in the control group had more than 1 episode of proven streptococcal infection associated with onset or flare of their chronic arthritis.

Table 2. Clinical and laboratory data on flares in the study group*
PatientJIA typeCulture/ASO titer
OnsetFlare 1Flare 2Flare 3Flare 4Flare 5
  • *

    NA = not available; GAS = group A beta hemolytic streptococcus; ASO = antistreptolysin O; SoJIA = systemic-onset juvenile idiopathic arthritis; see Table 1 for additional definitions.

2OligoNA/NAGAS/600 scarlet feverGAS/1,200GAS/NANA/965NA/1,130
4OligoGAS/NAGAS/800Only clinical tonsillitisGAS/404  
7OligoGAS/569 tonsillitisGAS/639    
8Ext. oligoGAS/334 tonsillitisGAS/NA tonsillitisGAS/408 tonsillitisGAS/400 tonsillitis  
9SoJIANA/NAGAS/NAGAS/267 tonsillitis   

Mean ± SD age at onset was 7.11 ± 3.88 years in the study group and 7.84 ± 4.91 years in the control group, with no statistically significant difference between the 2 groups. Concerning the type of JIA, 6 of the 9 patients in the study group had oligoarthritis (66%), 1 had extended oligoarthritis, 1 was polyarticular seronegative, none were seropositive, and 1 had systemic-onset JIA. In the control group, 13 (40%) of 32 patients had oligoarthritis, 5 had extended oligoarthritis, 6 had seronegative polyarthritis, 3 were seropositive, and 5 had systemic-onset JIA. The numbers were mostly too small for statistical analysis, but there were more patients with oligoarthritis in the study group and the difference was significant (P < 0.05). There was no difference between the study group and control group concerning positive ANA (4 of 9 and 16 of 32, respectively) and presence of uveitis (3 of 9 and 1 of 32, respectively).

The mean ± SD time in remission without treatment (except penicillin prophylaxis) was 3.1 ± 1.5 years for the study group and 1.8 ± 1.2 years for the control group, with a statistically significant difference between the 2 groups (P = 0.037). It is noteworthy that all patients in the study group were in remission with no treatment except penicillin prophylaxis whereas only 14 of the 32 patients in the control group were in remission without treatment at the time of the study.


The role of infection, particularly streptococcal infection, in the pathogenesis of autoimmune or autoinflammatory diseases has been investigated and discussed in many reports. There are some definite poststreptococcal diseases such as acute rheumatic fever, poststreptococcal reactive arthritis (9), and others where the causative role of streptococcal infection is less clear, such as Henoch-Schönlein purpura, psoriasis (4), uveitis (10), childhood polyarteritis nodosa (11), pediatric autoimmune neuropsychiatric disorders associated with streptococci, and more. Still, one has to be very careful in interpreting the results of uncontrolled trials. In a study reviewing the efficacy of antibiotic treatment or tonsillectomy in childhood psoriasis (4), only 1 controlled clinical trial was found, and this trial did not find a significant effect of antibiotic treatment on psoriasis.

The mechanism by which streptococcal infection may cause inflammatory diseases is believed to be based on molecular mimicry. In this model of molecular mimicry, self reactivity is triggered by cross-recognition of a self protein and an exogenous protein that bear the same sequence. It has been shown that serum from patients with acute rheumatic fever contains antibodies to an antigen in the membrane of the organism, the type 5 streptococcal M protein, which cross-reacts with myocardial tissue (12). A recent study demonstrated that human heart intralesional T cells react with myosin, valve-derived protein, and streptococcal M peptides (13).

The study by Massa et al (3) demonstrated that peripheral blood T cells of patients with juvenile DM lyse autologous target cells presenting an HLA class I binding peptide Myo (aa-114-122) derived from skeletal myosin. This suggests that this myosin-heavy chain epitope is a target autoantigen in juvenile DM. It has also been demonstrated that in cells from patients with juvenile DM with active disease, incubation with the streptococcal M5 (aa-367-375) peptide generated modest, although detectable cytotoxic activity against the human homolog Myo. Incubation of T cells with both peptides induced production of the proinflammatory cytokine tumor necrosis factor α in patients with juvenile DM but not in controls. These findings suggest that exposure to streptococcus encompassing homologous epitopes may trigger autoreactive T cells.

In view of these observations, it is also possible to assume that streptococci, especially those considered “arthritogenic,” may cause flare of chronic arthritis. In the present report, 9 patients with JIA, in whom onset or exacerbations were associated with streptococcal infections at least twice, are described and compared with all other patients with JIA diagnosed and treated at the same clinic during the same period. The 2 groups did not differ concerning age at onset, presence of positive ANA, and uveitis. There were more patients with oligoarthritis in the study group. There was a significant difference concerning time in remission without treatment (except penicillin), which was much longer for the study group.

Obviously, it should be asked whether or not these patients have a different disease such as poststreptococcal reactive arthritis, especially in the oligoarticular group. This possibility was considered, and in all the patients included there is no reasonable doubt concerning the diagnosis of JIA based on duration of arthritis, presence of uveitis, positive ANA, and response to treatment. This observation may lead us to the hypothesis that recognition of streptococcal infection as one of the factors causing flares in some patients with JIA and prophylactic treatment in these children may affect the course of their chronic disease.

There are several limitations to this study. This was an observational study with retrospective data collection with all the known problems of such. In contrast, the study was conducted in a single, primary center served by 1 pediatric rheumatologist, so at least the clinical approach and management were consistent and the patients were usually seen early after onset or flare. The study group was relatively small and some of the patients (mainly those diagnosed more recently) were started on penicillin prophylaxis following the second episode of flare associated with streptococcal infection. One could stipulate that in some cases this association is coincidental because streptococcal infection is very common in children and some are more prone to the infection than others. Taking all in consideration, this observation still deserves attention and a larger, possibly prospective study is warranted. In conclusion, recurrent streptococcal infection may cause worsening of the underlying disease in children with chronic arthritis, therefore it is worthwhile to look for this association and act upon it.


Dr. Barash had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Barash.

Acquisition of data. Goldzweig.

Analysis and interpretation of data. Barash, Goldzweig.

Manuscript preparation. Barash.

Statistical analysis. Barash.