Outbreak of Perianal Streptococcus Pyogenes Infection in a Day-Care Center

Streptococcus pyogenes or group A beta-hemolytic streptococcus (GAS) causes outbreaks of tonsillitis, scarlatina and skin infections such as impetigo. GAS can also cause a skin infection of the perianal region which has been called either perianal infection (PAI), perianal cellulitis or perianal dermatitis. Typical signs and symptoms of PAI, first described by Armen et al.,1 are perianal inflammation, itching, rectal pain and blood-streaked stools. The disease is an underrecognized entity and the diagnosis is often delayed or missed because similar symptoms and signs are commonly seen in other dermatologic conditions such as fungal and pinworm infections. In a report of 14 children the duration of signs and symptoms before diagnosis was 6 (range, 1 to 12) months.2 The infection responds well to oral penicillin therapy.1-3 In contrast to tonsillitis, scarlatina and impetigo, the epidemiology of streptococcal PAI is still largely unclear and only two reports of intrafamily spread have been published.3, 4 We describe here for the first time to our knowledge an outbreak of three documented patients (and four episodes) and two suspected cases of PAI in a day-care setting. Case history. The index case was a toilet-trained 3-year-old boy. He presented with a dark red, swollen eczema of the anal region. The eczema, which had been present for 1 week, was sharply demarcated from normal gluteal skin and it was itchy and irritated. His mother had noticed some blood on the stools. No impetigo or other skin abnormalities or systemic symptoms such as fever were present. A bacterial culture (Streptocult®, Orion Diagnostica, Espoo, Finland) of the skin around the anal sphincter was performed and heavy GAS growth was detected on the following day. The patient (Case 1) was treated with oral penicillin V (40 mg/kg/day divided in two doses for 10 days) and the symptoms and signs disappeared within 2 days. One week after the patient had completed the penicillin therapy the symptoms reappeared. The patient was reexamined at the same outpatient clinic and the signs and symptoms were identical with those seen during the first visit. Both S. pyogenes antigen test StrepA® (Abbott Laboratories, Abbott Park, IL) and anal streptococcal culture with bacitracin disk were positive for GAS. Oral cephalexin therapy was instituted and the recovery was again good. Methods. During the same visit the mother informed us that the boy attended a day-care center and that children in the same day-care group (n = 20) had had similar symptoms, i.e. itching eczemas of the anal region. We contacted the day-care center immediately and took both anal and pharyngeal cultures from all 20 children within 48 h. In addition a questionnaire concerning infections such as tonsillitis and PAI and the use of antimicrobials during the preceding 2 months was given to the guardians. Results. The cultures indicated (Table 1) that in addition to the index case (Case 1) two other children (Cases 2 and 3) had rich GAS growth in the anal region and clear symptoms of PAI. In addition to the positive anal culture Case 2 also harbored GAS in his throat. Furthermore the throat but not the anal culture of Case 6 (the sister of Case 3) was positive for GAS, but she was free of all symptoms. The questionnaires revealed that Cases 4 and 5 had a recent history of itchy eczemas in their anal regions. Because of these eczemas the children had visited their physicians 5 and 9 days before the day the cultures were obtained at the day-care center. Subsequently they were treated empirically with oral cephalexin because of to suspected perianal GAS infections. The recovery in both cases was rapid and uneventful. At the time of diagnosis of Cases 4 and 5 no cultures or other GAS tests were performed. One child (Case 7) had substantial group G streptococcal growth in both his buttock and throat but no symptoms.TABLE 1: Clinical and bacteriologic findings of the 20 day-care children The 5 cultured GAS isolates (Table 1) (3 anal isolates from Cases 1, 2 and 3 and 2 throat isolates from Cases 2 and 6) were first characterized serologically. The T serotyping and expression of the opacity factor5, 6 indicated that all 5 isolates were type T28 opacity factor-positive which is currently a very common serotype in Finland. (During 1995, 31% of superficial GAS isolates (n = 767) in the Helsinki area were of the T28 serotype (A Muotiala, unpublished data)). To test our hypothesis of an outbreak that was caused by a single pathogen more discriminatory characterizations were therefore required. The outbreak isolates and 11 geographically and chronologically related serotype T28 (opacity factorpositive) controls (pus (n = 6) and throat (n = 5) isolates) were first M-typed. The M types of all isolates were determined by immunodiffusion7 and all isolates expressed the R28 antigen. Next the isolates were genotyped by pulse field gel electrophoresis (PFGE)8 after endonuclease SmaI digestion. PFGE analysis is based on separation of enzymatically cleaved large genomic DNA fragments by electrophoresis during variable pulsing times. The PFGE gels indicated that the 5 outbreak isolates were indeed identical (Fig. 1) and that the 11 control isolates gave 8 different genotypic patterns, none of which was identical with the common pattern of the outbreak isolates.Fig. 1: SmaI macrorestriction profiles of the T28R28 isolates by PFGE. Lane 1, molecular weight markers (concatamers) of phage lambda; Lane 2, T28R28 type strain NCTC 8308; Lanes 3 to 7, outbreak isolates; Lanes 8 to 18, control isolates.Discussion. These data show that a small outbreak of three culture-confirmed perianal streptococcal infections and two additional nonverified clinical perianal infections occurred within 9 days in a group of 20 day-care children. The characterized PAI and the throat GAS isolates appeared to be genetically identical, suggesting that the infections were caused by a single S. pyogenes T28 clone. In addition to these cases we suspect that the two clinical noncultured cases were caused by the same GAS. The previous two reports on the epidemiology of PAI have demonstrated intrafamily spreading but not true outbreaks of such GAS infections. In 1970 Hirschfeld4 reported four cases of PAIs in two families. In the first family a 6-year-old girl and her 3-year-old brother had typical symptoms such as painful defecation and marked perianal erythema. These children had also had sore throats for ∼2 weeks before the onset of anal symptoms. In an unrelated family two siblings ages 5 and 3 years presented with identical history and symptoms. Bacterial cultures of all the four children showed anal GAS growth. One of these four children harbored GAS in the throat too. Further evaluation of the two families revealed GAS growth on the skin of three, in the anal region of one and in the throat of one additional member(s) of the family. In addition several family members had sore throats before the index cases presented with the anal symptoms. The epidemiology of these infections remained unclear. Fomites or bath water were, however, suspected as possible sources of the infection. In another report Kokx et al.3 describe 31 cases of PAIs caused by GAS in 19 families; 7 families had multiple cases of PAI. The family members were studied and intrafamily spread could be demonstrated in 50% of possible cases, but the disease was transmitted only to siblings and not to adults. They also serotyped the GAS strains isolated from throats and the perianal region and discovered that whenever positive, the pharyngeal T type was identical with that of the anal one. The most frequent typable T types detected from PAIs were T28 (n = 5), T4 (n = 4) and T2 (n = 1). The route of transmission could not be traced, but common bathing habits at home were suspected. The route of transmission in the outbreak described here could not be defined. Of the 20 children 2 harbored GAS in the throat, which may suggest that the transmission route was oral. The day-care center does not have a swimming pool and the children do not take baths together. However, this group of 20 toilet-trained children of approximately the same age (mean age, 4.2 years) have their own fairly isolated rooms and other facilities. They all used a single (miniature) toilet, which was not used by any other children of the day-care center or by the personnel. We therefore suspect that the transmission route was not oral but more likely the common toilet case and more precisely its GAS-contaminated plastic toilet seat. Unfortunately no bacterial cultures of the suspected toilet were performed because immediately after collection of the anal cultures the toilet and its cover were thoroughly cleansed. We have described for the first time a small outbreak of perianal streptococcal infection caused by genetically identical GAS. Although such cross-infections require ideal settings, i.e. intimate contacts like those in a day-care milieu where the pathogens can easily spread from child to child, these infections are probably more common than has been appreciated. Thus we believe that the described outbreak caused by this very common pathogen is not unique. This study also demonstrates the usefulness of the novel methods of molecular epidemiology. The excellent discriminatory power of the PFGE made it possible to distinguish the outbreak strain from the phenotypically identical control strains and thereby confirmed that a true outbreak had occurred. Acknowledgments. We thank Asha Tanna (Streptococcus and Diphtheria Reference Unit, Central Public Health Laboratory, London, UK) for technical assistance. Harri Saxén, M.D. Anna Muotiala, Ph.D. Timo Rostila, M.D. Jaana Vuopio-Varkila M.D.