416

Article

Vol. 11. No. 5

Eur. J. Clin. Microbiol. Infect. Dis., May 1992, p. 416--426 0934-9723/92/05 04.16-11 $3.00/0

Spectrum of Disease in Bacteraemic Patients during a Streptococcus pyogenes Serotype M-I Epidemic in Norway in 1988

A. Bucher 1., RR. Martin 2, E.A. H 0 i b y 2, A. Halstensen 3, A. ~ d e g a a r d 4, K.B. H e l i u m 5, L. Westlie 6, S. Hallan 7

All 87 known cases of bacteraemia due to Streptococcus pyogenes (beta-haemolytic group A streptococci) occurring during the peak of a nationwide outbreak in Norway (population 4.2 million) between January and June 1988 were reviewed. Clinical features varied widely and appeared largely to be dependent on the patients' age. The case fatality rate ranged from 11% in the age group under 30 years to 44 % in patients over 60 years. Clinical complications such as shock, severe renal or respiratory failure or serious local infection occurred particularly in 3040 59-year old individuals. Shock was manifest in 32 % of the patients and carried a 68 % case fatality rate. Chronic heart disease in the elderly and pneumonia seemed to be associated with a fatal outcome. In the 25 patients (29 %) who died the disease showed a fulminant course, 80 % dying within 48 hours after admission. However, 56 % of the patients had experienced symptoms for more than two days before admission, suggesting that early diagnosis and treatment might possibly have prevented the development of a serious disease. This study revealed a wide spectrum of clinical manifestations in bacteraemia cases in a unique epidemiological situation caused largely by a single serotype of Streptococcus pyogenes; 89 % of the 27 preserved bacteraemia strains carried the M-1 antigen. The observations call attention to the ability of these organisms to cause fuiminant clinical illness, indicating a probable increase in both invasiveness and toxicity of group A streptococci responsible for the epidemic.

Since the serological classification by Lancefield of beta-haemolytic streptococci, Streptococcus pyogenes (serogroup A beta-haemolytic streptococci, GAS) has been recognized as a significant human pathogen commonly associated with pharyngitis in children and cutaneous infections in individuals of all ages (1). In the pre-antibiotic era, GAS often caused serious invasive disease with a case fatality rate of up to 72 % (2). Subsequently, mortality from GAS infections 1Department of Infectious Diseases, UllcvSl University Hospital, N-0407 Oslo, Norway. 2Departments of InfectiousDisease Control and Bacteriology, National Institute of Public Health, N-0462 Oslo 4, Norway. 3Medical Department B, Haukeland University Hospital, N-5021 Bergen, Norway. 4Microbiological Laboratory, Molde County Hospital, N-6400 Molde, Norway. s University Hospital of Trondheim, N-7006 Trondheim, Norway. 60stf¢)ld Central Hospital, N-1600 Fredrikstad, Norway. 7Inhcrred Hospital, N-7600 Lcvanger, Norway.

decreased substantially, the decline starting before the introduction of effective antimicrobiat therapy (3, 4). A parallel decrease in the non-suppurative complications of streptococcal disease was observed. Serious infections caused by GAS, particularly bacteraemia, received little attention in the modern medical literature until the mid-1980s. Since then, there has been an increasing number of reports of fulminant infections with GAS bacteraemia and other serious clinical manifestations (5-17). A toxic shock-like syndrome associated with GAS has recently been reported (18-23) and a resurgence of acute rheumatic fever in the USA recorded (24). These events suggest that important changes in the pattern of GAS infections may still occur. In 1987-1988 a nationwide outbreak of localized and systemic GAS infections in Norway caused by serotype M-I organisms occurred. The number of bacteraemia cases reported in the Nor-

Vol. 11, 1992

wegian Notification System for Infectious Diseases increased threefold compared with earlier years (25, 26). In order to obtain more detailed information on the spectrum of clinical manifestations and the severity of the disease, we decided to review the records of patients reported with GAS bacteraemia in a six-month period during the peak of the outbreak.

Materials and Methods

417

serious local infection requiring major surgery; C) patients with moderately complicated illness, defined as serious local infection or moderate renal failure (creatinine 200--400 ~tmol/1) or more than three weeks of hospitalization due to the infection; D) patients with uncomplicated bacteraemia.

Bacteriological hlvestigations. GAS isolates were identified by standard methods and preserved for serotyping in 27 of the 87 patients included in the study. Typing of strains was preformed with T-sera from Chemapol, Czechoslovakia and tested for opacity factor (OF). Later the strains were M-typed by ourselves, while confirmatory M-typing was carried out by Dr. P. Kuzemenska, WHO Collaborative Laboratory for Streptococci, Prague, Czechoslovakia.

PatieJlts. GAS bacteraemia (defined as the finding of Streptococcuspyogenes in blood culture) must be notified in the form of a case report in Norway. Epidemiological information on cases of bacteraemia and other serious infections occurring during a GAS epidemic was collected a t the National Institute of Public Health in Oslo for the period from 1 January to 30 June 1988 (26). Retrospectively, medical and paediatric departments in Norway Were contacted by letter and sent a questionnaire with a request to give clinical data on all cases whether notified previously or not. All but three of the 84 known cases proved to fulfill the inclusion criteria and six new cases were recorded, giving a total of 87 cases which were analyzed in this study.

Collection of Clinical Data. We requested information on time of onset of symptoms, time of admission and specific previous medical events such as tonsillectomy, splcneetomy, frequent streptococcal infections, recent influenza or varicella infections, surgery or delivery. We also enquired about presenting symptoms and specific clinical findings on admission (temperature, blood pressure, mental function, cutaneous manifestations), as well as primary and secondary diagnoses, probable primary focus of infection, medical treatment prior to and after admission, and any underlying disease. In addition, data was requested on microbiological and laboratory results, clinical course and acute or late complications as well as Outcome.

Results Patient Data. A g e a n d g e n d e r of the 87 p a t i e n t s studied are shown in Figure 1. The median age was 48.2 years. There were 36 male and 41 female patients. Three of the six neonates were less than five days old at time of onset of the disease. Among the other children under the age of ten no particular age group predominated. Underlying Conditions. U n d e r l y i n g c o n d i t i o n s , which m a y have p r e d i s p o s e d to b a c t e r a e m i a a n d / or been of prognostic significance, are shown in Table 1. No underlying condition was found in 40 % (35/87) of the patients. Underlying conditions were clearly less frequent in patients < 60 years of age (16/51 = 30 %) compared to those > 60 years of age (30/36 = 83 %). In patients < 60 years of age the case fatality rate was 16 % (5/31)

Definitiot~s. We considered the infection to be comnaunity-acquired if a positive blood culture was obtained during the first 48 h after admission, oz' latex" if the patient presented clear evidence of infection at the time of admission. Neonates found to be bacteraemic while still in hospital were considered to have a hospital-acquired infection. Bacteraemia was considered to be cryptogenic if no obvious primary or secondary focus of infection was identified. A primary focus of infection was defined as a local clinical infection preceding bacteraemia. Infected sites that became apparent after bacteraemia was established and osteomyelitis, arthritis or meningitis were defined as secondary foci of infection.

P~aient Categories. The patients were subdivided into four categories (A-D) according to the severity of their clinical disease: A) patients who died; B) patients with severe clinical illness, defined as shock (systolic blood pressure < 90 mmHg), severe renal failure (serum creatininc > 400 ~tmol/l), treatment on a respirator or

18 16 14

i. Z

[]

male

6 4

< lrnont~

60

Age (years) Figure 1: Age and gender of 87 patients with group A streptococcal bactcraemia, January to June 1988, Norway.

418

E u r . J. Clin. M i c r o b i o l . I n f e c t . Dis.

Table 1: Underlying conditions and outcome in 87 patients with group A streptococcal baeteraemia related to age. Underlying condition

< 60 years (deaths)

> 60 years (deaths)

Total (deaths)

None

31(5)

4 (2)

35 (7)

One or more conditions

16 (4)

30 (13)

46 (17)

Malignancy" Chronic hearf disease Chronic venous insufficiency/lymphedema Diabetes mellitus Chronic alcoholism Immunosuppression b Rheumatic disease c Chronic liver disease Intravenous drugabuse Neonate

1 (0) 0(0) t (0) i (0) 3 (2) 1(0) 1 (0) 2 (0) 1 (0) 6 (2)

7 9 8 4 1 2 2

4(0)

2 (1)

6 (1)

51 (9)

36 (16)

87 (25)

Inadequate data availablc Total number of patients

(2) (6) (2) (2) (1) (1) (0)

8 9 9 5 4 3 3 2 1 6

(2) (6) (2) (2) (3) (1) (0) (0) (0) (2)

a7 carcinoma, 1 malignant lymphoma. b2 after renal transplantation, 1 after hypophysectomy with hormone substitution. c 2 rheumatoid arthritis, 1 polymyalgia rheumatica.

Table 2: Foci of infection in 87 patients with group A streptococcal bacteraemia.

Focus

Primary

Secondary

Skin Erysipelas Cellulitis Skin ulcer Other"

13 11 6 10

Respiratory tract Pharyngitis/tonsillitis Pneumonia b Sinusitis Other d

17 9 3 3

No known primary focus

15

Other Arthritis Osteomyelitis Meningitis Myositis Othel .e

-

6 4 3 3 4

Total t

87

23

3c -

a3 omphalitis, 2 varieella, 2 infected thrombophlebitis, 1 impetigo, 1 burns, 1 paronychia. b2 empyema. c primary focus: 2 pharyngitis, 1 erysipelas. d I dental abscess, 1 suppurative parotitis, 1 bacteremia after gastroscopy. e 1 neerotizing myometritis (puerperal), 1 necrotizing faseiitis, 1 retroperitoneal phlegmone, 1 ovarian abscess with peritonitis. f 9 eryptogenic bactcracmia cases (no known primary or secondary focus).

Vo1. t1,1992

419

for those without compared to 25 % (4/16) for those with underlying disease. Corresponding figures for patients > 60 years were 50 % (2/4) and 43 % (13/30). Six of nine patients with chronic heart disease, and three of four with chronic alcoholism died. No underlying condition was revealed in 28 % (7/25) of the patients who • died.

Focus of Infection. The primary focus of infection was defined clinically, but in 31 patients GAS were also isolated from the suspected focus (Table 2). Nine (5 male and 4 female) patients were considered to have had a cryptogenic bacteraemia: seven of these were over 60 years of age, one a neonate and one a 3-year-old child. Three patients with cryptogenic bacteraemia, alt > 60 years old, died. The most common primary focus of infection was the respiratory tract in the younger age group (48 % of patients under the age of 20 years vs. the skin in 30 %), and the skin in older patients (58 % of patients over the age of 60 years vs. the respiratory tract in 25 %) (Figure 2). Pharyngitis or tonsillitis as a primary focus of infection was found in 17 patients (19 %) and occurred in all age groups. Hospital-Acquired Infections. Eleven patients (13 %) had a hospital-acquired infection as follows: two patients from chronic care units, one who died without a known focus of infection and one who recovered from erysipelas; three neonates, one of whom developed meningitis; one patient with malignant lymphoma and several peripheral i.v. lines who underwent

70

60-

03 12n

~*

Sa %

40%

~

3o 30

"a ~

,1~

=o

%

20 %

E

10%

0 "P"" ,=1

3-4

4,7

)7

Onset.to-admission time (days)

Figure 3: Duration of symptoms before hospital admission and corresponding case fatality rate in patients with group A streptococcal bacteraemia, January to June 1988, Norway.

gastroscopy shortly prior to infection; one patient, who underwent gastroscopy prior to bacteraemia; one patient with a catheter-related peripheral thrombophtebitis; one patient with burns; one patient with post-operative pneumonia; and one patient with bullous pemphigoid and erysipelas. One case of puerperal necrotizing myometritis which although breaking out more than 48 h after admission, was considered to be community-acquired, because the patient had clinical tonsillitis at admission.

Onset-to-Admission Time. The duration of illness prior to hospital admission could be determined in 64 patients (Figure 3), 56 % (36/64) being symptomatic for more than two days, 27 % (17/64) for less than 24 h and 14 % (9/64) for more than seven days. The case fatality rate was 2836 % in patients with an onset-to-admission time of less than one week, as opposed to 44 % in those in whom the time was more than one week. Only a few patients had a history of previous streptococcal infections, tonsillectomy or recent influenza-like illness. Two patients developed bacteraemia after contracting varicella. Signs and Symptoms on Admission. Sixty-four

?.

"%°o%,°.o. 10 0

-"" not k'/lown °°°°,i~, ° .°"

I

~..20

I 20-40

I 4(I-60

I ~,r~O

Age (years)

Figure 2: Relationship between age and primary focus of infection in 87 patients with group A streptococcal bacteraemia, Janual2~ to June 1988, Norway.

percent (56/87) of the patients presented with signs and symptoms of local infection. Gastrointestinal symptoms were reported in 25 % (22/87) of cases: 11 had vomiting, seven diarrhoea and six nausea and abdominal pain. Specific information on temperature, blood pressure and mental function on admission was recorded for 81, 82 and 76 patients, respectively; 84 % (76/81) were febrile with a temperature of > 38 °C and

420

Eur. J. Clin. Microbiol. Infect. Dis.

•

~

Palechlae

Ecchymoses

bination of all three skin manifestations. Skin efflorescences were found in 50 % (10/20) of patients b e t w e e n 30 and 50 years, in 26 % (7/27) under 20 years of age and in 8.% (3/36) o v e r 60 years of age. Twenty-five p e r c e n t (5/20) died.

I - ~ ScaaaUnlf~m rash

4-

Z

t

I

< I month

400 pmol/i as inclusion criterion in category B), c o m p a r e d to only 10 % (3/29) of the remaining patients. Analyses to detect possible disseminated intravascular coagulopathy ( D I C ) were p e r f o r m e d infrequently. H o w e v e r , all positive results were found in categories A and B. T h e erythrocyte sedimentation rate ( E S R ) was only slightly elevated and showed little variation, 7 0 78 % of values being < 50 mm/h in all disease categories.

! 20-~I

Age

30-39

40-49

50-5g

,60

(years)

Figure 4: Relationship between types of lesion and age among 20 of the 87 group A streptococcal bacteraemic patients who developed skin manifestations (excluding skin infections). January to June 1988, Norway. 6 % (5/81) had a t e m p e r a t u r e of < 37 °C; 23 % (19/82) were hypotensive (systolic blood pressure < 90 m m H g ) . Mental confusion was described in 50 % (38/76) of the patients. Twenty-three percent (20/87) of the patients presented with skin lesions (Figure 4): seven had a scarlatiniform rash, three petechiae and three ecchymoses alone, while four had a c o m b i n a t i o n of scarlatiniform rash and petechiae, and one a com-

Table 3: Laboratory values in 87 patients with group A streptococcal bacteraemia within 24 hours of the first positive blood culture. (For disease categories A-D, see Materials and Methods.) N umber of patients* Parameter

Leucocyte ct)unt (x 109/1)

Thrombocyte count (x 109/1)

Values

> 15 10-15 3-10 t50

> 400 200--400 < 200

n=9

22

8

4

6

n = 63

n = 22

n = 12

n = 29

6 19 38

4 10 8

2 6 4

3 26

*n indic~ltes the number of patients in each disease category and in relation to the measured laboratory value.

Vol, 11, 1992

421

Course of Illness" and Outcome. The clinical COurse and manifestations of disease were in part dependent upon the age of the patient. Of the six neonates, two (33 %) died. Patients aged one month to 30 years had the lowest case fatality (5 %) and those over 60 years of age the highest (44 %). The overall case fatality rate was 29 %. The rate of clinical complications in surviving patients was highest in the age group 30 to 59 Years, where 33 % of the patients were classified in category B. Shock was recorded in 32 % (28/87), and 68 % (19/28) of these patients died. Four of the six other cases with a fatal outcome were reported to have a sudden death, assumed to be cardiac; two patients were dead on.admission. Acute renal failure with serum creatinine levels above 400 lamol/l was recorded in 30 % (26/87) of the patients; 37 % (3/11) of the surviving patients required haemodialysis. Respirator treatment was required in 16% (14187) of the cases, 7 1 % (10/14) of whom died. The available data were insufficient to assess the incidence of acute respiratory distress syndrome. Probable toxic shock-like syndrome was diagnosed in six patients with scarlatiniform skin rash (20, 21,23), three of whom died. All six had high fever, were mentally confused, and developed clinical shock and renal failure. Four patients had diarrhoea. Muscular pains, elevated liver and muscle enzymes in serum and thrombocytopenia Were also reported. Petechiae were seen in three

patients. Of the four patients between 30 and 40 years of age one had necrotizing myometritis, one suppurative parotitis with surrounding deep softtissue infection, one cellulitis and one no known focus of infection. One 62-year-old patient had pharyngitis and one 16-year-old patient an ovarian abscess and peritonitis. Suppurative complications were more common in children than in adults. There were three osteomyelitis, four arthritis, three meningitis and two empyema cases in patients under 20 years of age (12/26 = 46 %) but only one osteomyelitis and two arthritis cases in patients over 20 years of age (3/41 = 7 %). No local necrotizing infection in patients under the age of ten years nor any patients with endocarditis was seen. Surgical treatment due to serious local infection was carried out in seven patients. One patient with necrotizing fasciitis died after upper limb amputation. Fasciotomy was necessary in one case of myositis. Laparatomy was required in four cases: one hysterectomy for necrotizing myometritis, one oophorectomy for ovarian abscess, and drainage of one retroperitoneal phlegmonous infection and one intraabdominal, probably infected haematoma. Death occurred shortly after admission in most of the 25 fatal cases; 36 %died within 12 h, 64 % within 24 h and 80 % within 48 h. Time from first symptoms to death could be estimated in 19 patients; 42 % (9/19) died within three days and 3 1 % (6/19) after more than seven days. The case

Table 4: Local infections related to age in 25 patients who died of group A streptococcal bactcraemia. Age group

Infection

No. of patients

> 60 years (n = 16)

pneumonia erysipelas necrotizing fasciitis pharyngitis dental abscess none

5 5 1 1 1 3

40-49 years (n = 1)

suppurative parotitis

1

30-39 years (n =5)

pneumonia cellulitis pharyngitis

1 3 1

< 10 years (n = 1)

pneumonia with cmpyema

Neonatal (n = 2)

omphalitis none

1 1

422

fatality rate in all patients with pneumonia was 54 % (7/13) while five of six of patients over 60 years with pneumonia died. Table 4 shows the type of local infections in cases with a fatal outcome. Twenty-eight percent (4/11) of patients using nonsteroidal anti-inflammatory drugs prior to admission died, all being older than 60 years, giving a case fatality rate of 67 % (4/6) for this age group using such drugs.

Treatment. Detailed information on antibiotic treatment was recorded for 94 % (82/87) of the patients. Four patients who died were not treated with antibiotics. Seventy-three patients were treated initially with a beta-lactam antibiotic, often in combination with an aminoglycoside; treatment was started in the majority of the Patients shortly after admission. Four patients received erythromycin and one patient clindamycin. Bacteriological Results. GAS isolates from 27 patients were preserved for serotyping: 89 % (24/27) carried the M-1 antigen (4 were T-nontypable and 20 T-l; all opacity factor (OF) negative). Two isolates were M-3, R-3, both T-nontypable and OF negative; the corresponding patients both developed shock and one of them died. The last strain was M-non-typable (T-4, OF negative), All M-1 and M-3 strains were markedly mucoid (26). Discussion

Keefer et al. (2) described the clinical picture of GAS bacteraemia in 246 patients in the pre-antibiotic era. Studies published during recent decades (5-16) have been of sporadic cases, often occurring over several years or in selected patients. In late 1987 the Norwegian Notification System for Infectious Diseases started to record an increasing number of serious GAS infections (25), Subsequently it became clear that a nationwide outbreak of GAS infections, including many fulminant cases, was emerging with a peak incidence in the first six months of 1988 (26). Recent reports from the UK, other Nordic countries and North America (27-31) have also indicated a reappearance of severe GAS infections, including many bacteraemia cases, and dominance of serotype M-1. The epidemiological situation in Norway during 1987-1988 enabled us to study a series of bacteraemia cases caused by serotype M-1 organisms

Eur. J. Clin. Microbiol. Infect. Dis.

(26), which recent investigations have shown to be of a slightly different genetic clone of serotype M-1 than that found in earlier Norwegian strain collections (D.A. Caugant et al., unpublished data). The data on 87 bacteraemia cases analyzed retrospectively in this study, although collected over only six months in a population of 4.2 million, is to our knowledge the largest collection of clinical data on GAS bacteraemia reported since the study of Keefer et al. (2). While Keefer et al. (2) found GAS bacteraemia cases in all age groups, this disease has been reported in more recent times mainly in elderly patients (many with underlying conditions) and infrequently in younger patients with underlying conditions such as leukemia or intravenous drug abuse (14, 15, 26, 32). The formerly feared neonatal GAS bacteraemia has only been reported a few times in the modern medical literature (33, 34). In our material all age groups were represented with no clear differences between sexes (Figure 1), and in contrast to Keefer et al. (2), who reported many cases of puerperal infection, we registered only one such case. The frequency of GAS bacteraemia in adolescent and middle-aged patients in our series was remarkably high and clearly different from that recorded in previous years in Norway (26). Our rate of 13 % hospital-acquired cases is comparable to that reported in other recent studies (7-20 %) (5, 6, 8, 9, 12) although nosocomial acquisition was as high as 47 % in a somewhat more selected patient material (35). The primary focus of infection could be recognized in the majority of our cases (Figure 2 and Table 2). Similarly to Keefer et al. (2) we found a clear predominance of the respiratory tract as the primary focus of infection in younger patients, whilst bacteraemia most often originated in the skin in the elderly. Although the association between GAS bacteraemia and throat infection has been considered unusual in modern times (36), 19 % of our patients had pharyngitis or tonsillitis as the primary focus of infection (Table 2). Cryptogenic GAS bacteraemia has been considered unusual and associated with a relatively high case fatality in many studies (5, 6, 8), although Dan et al. (9) reported 31% cryptogenic cases in their series. We found cryptogenic bacteraemia in 10 % of our patients (9/87); 33 % of these died compared to 25 % of non-cryptogenic cases. Underlying conditions that may have been predisposing factors were identified more fre-

Vol. 11, 1992

423

n=28

quently in the elderly over 60 years of age than in those under 60 years. Other investigators have found a similar distribution of underlying conditions in the different age groups (5). Our findings give no clear indication as to whether an underlying disease or perhaps simply the age of the patient should be considered a predisposing factor. However, they do show that virulent strains of GAS are able to cause serious bacteraemia in previously healthy young individuals. Patients with Underlying conditions did not have a higher case fatality rate in general but old age, chronic heart disease and chronic alcoholism seemed to be unfavorable prognostic factors. The onset-to-admission time varied substantially (Figure 3). Symptoms of local infection preceded development of bacteraemia by more than two clays in over half of the patients. Although not clearly confirmed by our results, this suggests that early diagnosis and treatment might have prevented development of serious illness. The presenting signs and symptoms in our patients Were comparable to those reported in previous studies. Four of five patients with hypothermia died. Many patients presented with gastrointestinal disturbances. Such symptoms are frequently also found in cryptogenic cases (6, 8, 14), sometimes mimicking an acute abdomen or acute gastroenteritis and creating a pitfall in the initial diagnostic procedure, A striking finding was that about a quarter of our Patients presented with petechiae, ecchymoses and/or scarlatiniform skin rash (Figure 4). These signs may have been underreported in our material, although petechiae and ecchymoses, Particularly in the elderly, may have causes other than GAS bacteraemia. There are relatively few reports in the literature of petechiae in patients with GAS bacteraemia (10, 13, 21, 37). Scarlatin{form rash in patients with GAS infections has been thought to be caused by previous sentization to pyrogenic exotoxins (21, 38) but it remains unclear which mechanism is responsible for the development of skin haemorrhages, or for the fact that patients over 60 years of age had no Petechiae and rarely scarlatiniform rash in our Study. We found no obvious association between skin haemorrhages, thrombocytopenia and disSeminated intravascular coagulation. Certain laboratory values on admission were associated with a severe course of disease (leucocyte count < 10 x 10611, thrombocyte count < 100 x 106/I, CRP > 150 rag/l, and creatinine 200 lamolll). These parameters may therefore

n=23

n=3B

100.

80.

60.

40.

g

20.

0, 60 A g e (years)

cale~ory C {moderately severe disease) Category B {severe disease) Category A (lalal cases)

Figure 5: Sevcrity of disease related to age in group A streptococcal bacteracmia, January to June 1988, Norway.

have prognostic importance in the initial clinical assessment of patients with GAS bacteraemia. The variation in clinical manifestations with age was remarkable. Suppurative metastatic complications were considerably more frequent in children and adolescents than in adults. Two neonates developed meningitis, an uncommon manifestation of GAS infection (39), and one 18year-old patient developed meningitis secondary to sinusitis. Empyema is a well known complication of GAS pneumonia occurring particularly in children (40). Two patients, both under 20 years of age, developed empyema and one of them died, whilst several other cases of GAS pneumonia and empyema without confirmed bacteraemia were reported in Norway in the same period (26). The low incidence of underlying disease, relatively low fatality rate and high rate of suppurative complications are similar to that observed in other recent studies of GAS bacteraemia in children (10, 11), although the rate of suppurative complications is even higher in our material. We have no good explanation for the notably high rate of severe clinical complications in the 30-59 years age group (51% of the patients belonged to category A or B, Figure 5). However, premorbid acquired immunity to streptococcal virulence factors and other host factors may be important determinants of the clinical course of the GAS bacteraemia. It is also possible that some benign cases of bacteraemia in middle-aged persons were treated ambulatorily while such cases more likely resulted in hospitalization in the other age groups.

424

The case fatality rate was particularly high in the elderly and the clinical course rapid, despite the fact that 40 % of the patients > 60 years of age had an uncomplicated disease mostly associated with erysipelas or cellulitis (Figure 5). The four sudden deaths in this age group were probably of cardiac origin and streptococcal toxins may have been involved. There was a possible relationship between a higher case fatality rate in the elderly and use of nonsteroidal anti-inflammatory drugs prior to onset of GAS disease. Such drugs have recently been associated with a particularly fulminant form of streptococcal soft tissue infection (41,42). Two different disease entities caused by GAS deserve special mention: the streptococcal toxic shock-like syndrome (STSLS) and soft tissue infections. STSLS, which may also occur without bacteraemia, has only recently been recognized (21, 23, 43) and a clear case definition has not been established. We classified six cases with scartatiniform skin rash as probable STSLS but this may have been an underestimate of the real number as our study was not designed to identify patients with this condition. Half of these patients died. Fourof the patients were in the 30 to 59 year age group. STSLS has mostly been associated with deep necrotizing soft tissue or genital infections but has also been descrY,bed in patients with pharyngitis or, as in two of our patients, without any known portal of entry (21, 23). Skin and soft tissue infections due to GAS occur in several forms. Clinically it may be difficult to differentiate erysipelas from cellulitis and cellulitis from necrotizing fasciitis or myositis. We used the diagnoses provided by the participating hospitals. All cases of necrotizing fasciitis and myositis were verified histologically or by tissue culture. Middle-aged and elderly patients had deep soft tissue or necrotizing GAS infections more frequently than younger patients; the only necrotizing infection in a child was a histologically verified myositis related to suppurative arthritis in a 12year-old girl. In addition to the cases included in this study, other cases of severe necrotizing soft tissue infections caused by serotype M-1 but without confirmed bacteraemia occurred in the study period.

Eur. J. Clin. Microbiol. Infect. Dis.

The course of the disease after recovery from the acute illness was favorable in most patients. Renal failure seemed to be reversible within a few weeks, although a number of patients with serious local infections suffered significant local sequelae. The lack of reports of non-suppurative complications is interesting. The relative importance of toxaemia and invasiveness in the pathogenesis of these bacteraemia cases remains unclear. Some clinical features support the hypothesis that streptococcal toxins play an important pathogenic role. Toxic shock-like illness, scarlatiniform rash, necrotizing local infections, sudden cardiac death and multiple organ failure may represent toxin-mediated clinical phenomena. However, many cases with minor primary foci of infection usually not associated with bacteraemia, such as throat infections, cryptogenic cases and serious suppurative infections suggest increased invasiveness of the GAS strains. So far little is known about the relative contribution of individual streptococcal toxic products, such as pyrogenic and cytolytic toxins, or toxic cellular constituents, such as peptidoglycan, to the clinical traits of GAS bacteraemia (44). Investigations in Sweden suggest that the Scandinavian GAS strains produce pyrogenic exotoxin type B (45 and S. Holm, unpublished data) in contrast to strains isolated recently in the USA, which produced type A exotoxin (21). An important determinant of virulence is also the Mprotein which inhibits phagocytosis of streptococci. Most GAS strains in this material were of serotype M-1 and had a striking mucoid appearance (26). Knowledge of current changes in the epidemiology of GAS infections and the actual M-types involved is relevant also for primary care physicians. When invasive and virulent GAS strains predominate, early and aggressive treatment of probable GAS infections appears at present to be the only alternative in preventing development of serious disease with a possibly fatal outcome.

Ackn owledgements Almost two-thirds of deaths occurred within 24 h after admission, a somewhat higher proportion than that reported in recent studies in the UK (5, 6). Most of these patients received adequate antibiotic treatment after admission without any obvious effect on the course of the disease at this stage.

We would like to thank all clinical colleagues in Norway who enabled us to collect the data on which this paper is based. We sincerely lhank Professor J.N. Bruun and Professor A. Lystad for valuable discussionsand support. We also want to thank Dr. P. Kuzcmenska, Prague, Czechoslovakia who carried out the confirmatory Mtyping of isolates.

Vol. 11,1992

References 1. Lancefield RC: A serologic differentiation of human and other groups of hemolytic streptococci. Journal of Experimental Medicine 1933, 57: 571-595. 2. Keefer CS, lngelfinger FJ, Spink WW: Significance of hemolytic streptococci bacteremia: a study of 246 patients. Archives of Internal Medicine 1937, 60: 10841097. 3. Madsen ST:, Scarlet fever and erysipelas in Norway during the last 100 years. Infection 1973, 1: 76-81. 4. Quinn RW: Epidemiology of group A streptococcal infections: their changing frequency and severity. Yale Journal of Biology and Medicine 1982, 55: 265-270. 5. lspahani P, Donald FE, Aveline AJD: Streptococcus pyogenes bacteraemia: an old enemy subdued but not defeated. Journal of Infection 1988, 16: 37--46. 6. Francis J, Warren RE: Streptococcus pyogenes bacteraemia in Cambridge. A review of 67 episodes. Quarterly Journal of Medicine 1988, 256: 603--613. 7. Centers for Disease Control: Group A beta-hemolytic streptococcal bacteremia-Colorado 1989. Morbidity and Mortality Weekly Report 1990, 39: 3-6. 8. Bibler MR, Rouan GW: Cryptogenic group A streptococcal bacteremia: experience at an urban general hospital and review of the literature. Reviews of Infectious Diseases 1986, 8: 941-951. 9. Dan M, Maximora S, Siegman-lgra Y, Gutman R, Rotmensch HH: Varied presentations of sporadic group A streptococcal baeteremia: clinical experience and attempt at classification. Reviews of Infectious Diseases 1990, 12: 537-542. 10. Wong VK, Wright HT: Group A beta-hemolytic streptococci as a cause of bacteremia in children. American Journal of Diseases of Children 1988, 142: 831-833. 11. Christie CDC, Havens PL, Shapiro ED: Bacteremia with group A streptococci in childhood. American Journal of Diseases of Children 1988, 142: 559-561. 12. Dubey B, Francioll P, Poii S, Glauser MP: Septicemies a streptocoques du groupe A (S. pyogenes). Schweizerische Medizinische Wochenschrift 1983, 113:41-44, 13. Enzenberger R, Helm ER, Stille W: A-Streptokokken-Septik~imien. Analyse von 18 F~illen bei Erwachsenen. Deutsche Medizinische Wochenschrift 1984, 109: 899-902. 14. Borg NL, Kish MA, Kauffman CA, Supena RB: Group A streptococcal bacteremia in intravenous drug abusers. American Journal of Medicine 1985, 76: 569-

574. 15. Lentnek AL, Giger O, O'Rourke E: Group A betahemolytic streptococcal bacteremia and intravenous substance abuse. A growing clinical problem? Archives of Internal Medicine 1990, 150: 89-93. 16. Barnham M: Invasive streptococcal infections in the era before the acquired immune deficiency syndrome: a 10 years' compilation of patients with streptococcal bacteremia in North Yorkshire. Journal of Infection 1989, 18: 231-248. 17. Maclennan AC: Invasive streptococci. Lancet 1990, i: 109. 18. Begovac J, Gwajnicki B, Kuzmanovic N: Invasive streptococci. Lancet 1990, i: 109. 19. Cone LA, Woodard DR, Schlievert PM, Tomory GS: Clinical and bacteriologic observations of a toxic shock-like syndrome due to Streptococcus pyogenes. New England Journal of Medicine 1987, 317: 146-149.

425

20. Barrier T, Dascal A, Carroll K, Curley FJ: "Toxic strep syndrome". A manifestation of group A streptococcal infection. Archives of Internal Medicine 1988, 148: 1421-1424. 21. Stevens DL, Tanner MH, Winship J, Swarts R, Ries KM, Schlievert PM, Kaplan EL: Severe group A streptococcal infections associated with a toxic shock-like syndrome and scarlet fever toxin A. New England Journal of Medicine 1989, 321: 1-7, 22. Conolly TJ, Pavelka D J, Lanspa EF, Conolly TL: Toxic shock-like syndrome associated with necrotizing Streptococcus pyogenes infection. Henry Ford Hospital Medical Journal 1989, 37: 69-72. 23. Kfhler W: Streptococcal toxic shock syndrome. Zentralblatt fiir Bakteriologie 1990, 272: 257-264. 24. Veasy LG, Wiedmeler SE, Orsmond GS, Ruttenberg HD, Boucek MM, Roth SJ, Tail VF, Thompson JA, Daly JA, Kaplan EL, Hill HR'. Resurgence of acute rheumatic fever in the intermountain area of the United States. New England Journal of Medicine 1987, 316: 421--427. 25. Marlin PR, Haihy EA: Streptokokkinfeksjoner i Norge 1987 og 1988. Oslo: National Institute of Public Health, Meldesystem for Infeksjonssykdommer 1988, 16: week 10. 26. Marlin PR, H~iby EA" Streptococcal serogroup A epidemic in Norway 1987-1988. Scandinavian Journal of Infectious Diseases 1990, 22: 421-429. 27. Gaworzewska E, Colman G: Changes in the pattern of infection caused by Streptococcus pyogenes. Epidemioiogy and Infection 1988, 100: 257-269. 28. Holm S: Serious infections caused by group A streptococci. Stockholm: Statens Bakteriologiska Laboratorium, Epid Aktuellt 1988, I1: week 47. 29. Henrichsen J, Hoffman S, Rothgardt NP, Rtmne T: Infection with group A streptococci. Copenhagen: Statens Seruminstitut, Epi-Nyt 1988, week 50. 30. Demers B, Simor AE, Vellend H, Low DE: Severe group A streptococcal disease, Toronto, Ontario. Canada Diseases Weekly Report 1990,16-5 h 259-260. 31. Schwartz B, Facklam RR, Breiman RF: Changing epidemiology of group A streptococcal infection in the USA. Lancet 1990, ii: 1167-1171. 32. Dudding B, Humphrey GB, Nesbit ME: Betahemolytic streptococcal septicemias in childhood leukemias. Pediatrics 1969, 43: 359-364. 33. Coulter JBS, Buchannon CR, Vellodi A, Hart CA, Sills JA: Group A streptococcal infection in the newborn. Lancet 1984, ii: 1473-1474. 34. Lehtonen O-P, Ruuskanen O, Kero P, Hollo O, Erkkola R, Salmi T: Group A streptococcal infection in the newborn. Lancet 1984, ii: 355-356. 35. Roberts FJ: Group A and group B l~-hemolytic streptococcal bacteremia. Reviews of Infectious Diseases 1988, 10: 228-229. 36. Barnham M: Bacteraemia in streptococcal infections of the throat. Journal of Infection 1983, 7: 203-209. 37. Conner WT: Fulminant Streptococcus pyogenes infection. British Medical Journal 1981, 282: 651. 38. Schlieverl PM, Bellin KM, Watson DW: Reinterpretation of the Dick test: role of group A streptococcal pyrogenie exotoxin. Infection and Immunity 1979, 26: 467-472. 39. Murphy DJ: Group A streptococcal meningitis. Pediatrics 1983, 71: 1-5.

426

40. Molteni RA: Group A beta-hemolytic streptococcal pneumonia. Clinical course and complications of management. American Journal of Diseases of Children 1977, 131: 1366-1371. 4t. Rimaiiho A, Riou B, Richard C, Auzepy P: Fulminant necrotizing fasciitis and nonsteroidal anti-inflammatory drugs. Journal of Infectious Diseases 1987,155: t43-146. 42. Frankish PD, Mason GH, Allen PR, Milsam FP, Chrislmas Th Acute streptococcal necrotising fasciitis. New Zealand Medical Journal 1988, 101: 625--626.

E u r . J. Clin. M i c r o b i o l . Infect. Dis.

43. Hribalova V: Streptococcus pyogenes and toxic shock syndrome. Annals of Internal Medicine 1988, 108: 772. 44. Wannamaker LW: Streptococcal toxins. Reviews of Infectious Diseases 1983, 5, Supplement 4: 723-732.

45. Slegmayr B, Burman LA,, B[ickiund U, Holm S, Setlergren B, Svanslr6m M: Tvh fall av allvarlig grupp A-streptokockinfektion. L~ikartidningen 1989, 86: 3672-3675.