Severe invasive streptococcal infections: Since enactment of the Infectious Diseases Control Law, 302 cases of TSLS have been reported between April 1999 and September 3, 2004, under the National Epidemiological Surveillance of Infectious Diseases (NESID); death was confirmed in 114 cases by the time of notification, of which 65 (60%) died within three days. Twenty-two cases (10 deaths) were reported in 1999 (April to December), 46 cases (19 deaths) in 2000, 48 cases (19 deaths) in 2001, 89 cases (33 deaths) in 2002, 56 cases (15 deaths) in 2003, and 41 cases (18 deaths) in 2004 (as of September 3). No particular seasonal characteristics in disease incidence were observed (Fig. 1). Cases were distributed throughout 46 prefectures (with the exception of Iwate Prefecture) without any demonstrable variations by region (Fig. 2). Many cases were in the 60-year age group, with the average age being 55.8 years. The sex ratio was 1.42 (177 males, 125 females), with more males in the 40-, 50-, and 60-year age groups (Fig. 3).
Etiological agents were recorded in only approximately 50% of the cases, of which 150 were group A, two were group C, six were group G, and 10 were not identified. These results show that TSLS is not always caused by group A streptococci (see p. 257 of this issue).
Although serotyping of the M protein, the pathogenic factor of group A streptococci, is important, such serotyping is difficult to perform. Therefore, genotyping of the M protein gene (emm ) has been utilized in recent years. In contrast, T serotyping is relatively easier to perform than M serotyping, allowing many laboratories to utilize this method as a means of conducting epidemiologic investigations (see p. 254 & 256 of this issue). Since the first report of typical TSLS cases in Japan in 1992, infectious agent surveillance has been conducted at the Streptococcus Reference Center of the Associations of Public Health Laboratories for Microbiological Technology (see IASR, Vol. 18, No. 2). Of 109 cases in which T serotyping and emm genotyping were conducted between 2000 and August 2004, 46 were typed as T1/emm1 , accounting for 42% of all cases. Ten cases were typed as T3/emm3 , nine as TB3264/emm89 , eight as T28/emm28 , seven as T12/emm12 , three each as T4/emm4 , T6/emm6 , and T22/emm22 , and one as TB3246/emm103 (Fig. 4). Of the increasing untypable T serotypes, seven were emm49 , three each were emm58 , emm75 , and emm81 , and one each was emm73 , emm77 , and emm78 . Unlike strains derived from patients with pharyngitis, as described later, TSLS-derived strains are frequently characterized as T1/emm1 , although the number of isolates of this type has been low in 2004 (see p. 254 of this issue). Untypable T/emm49 had not been isolated prior to 1999 (see IASR, Vol. 24, No. 5) (see the report of the Streptococcus Reference Center, http://idsc.nih.go.jp/pathogen/refer/str2003f.pdf).
Group A streptococcal pharyingitis: Numbers of cases of group A streptococcal pharyngitis per sentinel clinic [( ) denotes total reported cases] reported under the NESID are as follows: 53.10 (158,143) in 2000, 51.32 (154,932) in 2001, 51.38 (155,999) in 2002, 54.73 (166,437) in 2003, and 53.05 (161,115) in 2004 (as of week 38). Cases reported per sentinel clinic have been declining every summer (Fig. 5); most are between 4 to 7 years of age; no noticeable changes in age distribution (see IASR, Vol. 21, No. 11) were seen.
From 2000 to 2003, there were between 1,909 and 2,627 yearly reports of group A Streptococcus isolations in which T serotyping was conducted at prefectural and municipal public health institutes (PHIs). T serotypes most frequently isolated every year were T12, T1 and T4, with these three serotypes accounting for more than 50% of all isolates (Fig. 4). The proportion of serotypes T12, T1, T4, T11, T28 and TB3246, fluctuated little year to year. In contrast, drastic increases in serotypes T3, T6, and T18 were observed, with peaks in 1993-1994 and 2002, 1997, and 1994, respectively. All three serotypes exhibited large yearly fluctuations in isolation frequency (see http://idsc.nih.go.jp/iasr/prompt/circle-g/past/st5.gif). Monthly changes of T serotype distributions after 2000 depict increases in each serotype in different seasons, although simultaneous decreases in the summer were evident (Fig. 6).
In addition to person-to-person infections, foodborne outbreaks of pharyngitis have also been reported in recent years (IASR, Vol. 18, Nos. 11 and 12; Vol. 19, No. 12; Vol. 20, No. 5; and Vol. 25, Nos. 2 and 4). Attention therefore must be paid to pharyingeal symptoms in food handlers and other kitchen staff.
Conclusion: Reporting criteria of TSLS under the Infectious Diseases Control Law is based on diagnostic criteria proposed by CDC, USA in 1993, with the pathogenic agent being restricted to group A streptococci. However, according to recent findings, isolations of group C and G streptococci from TSLS cases have been increasing, thereby necessitating revisions in TSLS reporting criteria.
Furthermore, under the current system, if a death due to TSLS occurs after reporting, a revised report is requested because the outcome is often not known at the time of the initial case report. Hereafter, a systematic method of verifying information on case outcome or prognosis will be required.
As a result of drug susceptibility testing performed at the aforementioned reference center, strains resistant to clindamycin, one of the first-line drugs for streptococcal infections, have been reported after 2000 (see p. 254 of this issue). However, because TSLS-derived strains examined at the reference center have accounted for only 40% of the reported cases under the NESID, it will be necessary to strengthen the system of collecting case strains in order to better understand the epidemiology of streptococcal infections in Japan. It will be important to isolate the etiological agent from patients with TSLS and group A streptococcal pharyngitis, monitor trends in serotypes and drug susceptibility, and feedback accurate information for the early diagnosis and treatment of the patients.