Erythromycin-resistant group G streptococci in an isolated northern Canadian community

The susceptibility of groups A. C. and G streptococci isolated from pharynx or s kjn in two north ern Canadian native communHies during a one year study of the epidemiology of streptococcal infec tion was determined for penicillin. erythromycin a nd clindamycin using an agar dilution method. Organisms studied included 725 group A. 82 group C. and 184 group G streptococci. All organisms were susceptible to penicillin (minimum inhibitory concentration (MIC) range less than 0.004 to 0.015 J.lg/mL: M!Cgo 0.015 JJg/ mL) a nd clindamycin (range 0.007 to 0.06 JJg/mL: MICgo 0.06 J.lg/mL) with no differences observed between s treptococcaJ groups. For erythromycin. groups A and C were generally susceptible (range less than 0.007 to 0.030 J.lg/mL: MlCgo 0.03 J.lg/ mL: and range 0.007 to 1.0 J.lg/mL: MlCgo 0.06 J.lg/mL. respectively). GROUP G STREPTOCOCCI MAY BE ISOLATED AS PART of lhe normal flora of lhe pharynx, intestine, vagina and skin. with symptomatic infection including pharyngitis and invasive disease ocSection of lnfect.iJ:Ju.s Diseases. Depanment qf Medicine; Norlhem Medical UniL Department of Community Heallll Sciences: Departments of Pedialrics and Medical Microbiology. U11Iversily of Manitoba: and Public Healtlll...aboratcxy Seroice. London. UK Correspondence and reprints: Dr LE NicoLLe. Health Sciences Centre. MS675D · 820 Sherbrook Street. Winnipeg. Manitoba R3A 1 R9. Telephone (204) 787-4655 Receiuedfor publication January B. 1990. Accepted February 13. 1990 CAN J INFECT DIS VOL l NO l SPRING 1990 Group G was less susceptible (range 0.007 to grea ter tha n 2.0 J.lg/mL: MlCgo greater than 2.0 J.lg/ mL) with 38% of a ll isolates having an MIC greater than or equal to 1 JJg/mL. On review of group G isolates. 100 of 100 from one community were susceptible (MIC less than 0.007 to 0.03 ).lg/mL) and 73 (87%) of 84 from the second community were resistant. All res istant strains tested were type T 16. These data suggest that erythromycin-res istant group G s treptococci may occur with high prevalence in certain populations and tha t pattem s of anlimicrobiaJ susceptibility in isolated communities may be highly community-specific . Can J Infect Dis 1990; 1(1):3-6

Group G was less susceptible (range 0.007 to grea ter tha n 2.0 J.lg/m L: MlCgo greater than 2.0 J.lg/ mL) with 38% of a ll isolates having an MIC greater than or equal to 1 JJg/mL.On review of g roup G isolates.100 of 100 from one community were susceptible (MIC less than 0.007 to 0.03 ).lg/mL) and 73 (87%) of 84 from the second community were resistant.All res istant strains tested were type T 16.These data suggest that erythromycin-res istant group G s treptococci may occur with high prevalence in certain populations and tha t pattem s of anlimicrobiaJ susceptibility in isolated communities may be highly community-specific .Can J casion a lly iden tified (l-3).These organisms a re more frequently tolerant of antibiotics than other beta-hemolytic streptococci (4,5), and plasmidmediated resistance lo erythromycin has been described (6).Erythromycin resistance a mong group A streptococci is well documented (3. 7 .8). and in previou s reports.a high prevalence of erythromycin-resistant group A streptococci has been reported in certain geographic areas (7).In this repor t lhe antimicrobial s u sceptibili ties of beta-hemolytic streptococci are described in two isolated northern Canadian commu nities, one of which h ad a h igh prevalence of group G streptococci resistant to erythromycin.

Study populations:
A prospective study of group A streptococcal carriage and d isease was undertaken in two northen1 Canadian native communities from November 1984to October 1985 (9).The communities were one Inuit and one native Indian community.both with populations of approximately 1200.Both com munities are geograph ically isolated in northern Canada with convenient access by air only.The study design included three prevalence surveys of pharyngeal beta-hemolytic streptococcal carriage in school children in November 1984 and Febmary and May 1985.a nd a prospective study of the incidence of group A streptococcal pharyngitis and impetigo with pharyngeal and skin swabs from symptomatic residents throughout the study year.Microbiol ogic methods: Pharyngeal and s kin swabs were inoculated and incubated.and organisms were isolated and identified in the communily for the prevalence surveys.For the incidence study.specimens were forwarded by air to the study centre in Winnipeg.Manitoba for processing.All specimens were inoculated onto Mueller-Hinton agar layered with sheep blood agar and incubated overnight at 37°C in room air.Gram-positive.catalase-negative cocci producing bela-hemolytic colonies were presumptively identified as beta-hemolytic streptococci.Initial grouping of streptococci was done using a CAMP plate with a bacitracin A and SXT disc.and subsequently confirmed using the Streptcx Latex Kit (Wellcome Diagnostics).T-typing of group G streptococci was performed at the Pub lic Health Laboratory in London.England using previously described methods ( 1 0).
Antimicrobial s usceptibility testing used an agar dilution method (11).Mueller-llinton agar was supplemented with 5% defibrinated sheep b lood agar.Organisms were grown overnight in Todd Hewitt broth and d il uted to 10 8 colony forming units (cfu)/mL in Mueller-Hinton broth using a McParland standard.A 1:10 dilution in broth to 10 7 cfu/ mL was made.and 10 4 cfu/mL was applied with a Ste~rs replicator to b lood agar plates of twofold increasing antibiotic concentration.Plates were examined after 24 h of incubation at 37°C in room air.The minimal inhibitory concentration (MlC) was the lowest antimicrobial concentration which prevented any growth of the organism.

RESULTS
1\ total of 725 group A. 82 group C and 184 group G streptococci were isolated.The MICs of the three beta-hemolytic streptococci to penicillin G. erythromycin.and clindamycin are shown in F'igurc 1.The MICs were generally comparable for penicillin and clindamycin.F'or erythromycin.groups A and C streptococci had comparable MICs.but group G s treptococci demonstrated a biphasic distribution.
The group G streptococci with e levated M!Cs were a ll isolated from residents of the native Indian communily (F'igurc 2).All isolates of group G s treptococci from the Inuit commun ity were susceptible lo erythromycin.whereas 86% of isolates from the native Indian community had MICs of at least I )lg/mL.The MICs of group G streptococci for the three prevalence surveys in the native Indian community were greater than or equal to l )lg/mL for 24 of 29 (83%) isolates at the first s urvey.21 of 22 (95%) at the second survey and 16 of 16 at the third survey.Thus. the proportion of group G streptococci with elevated MICs increased consecutively throughout the study  period.Twelve (71 %) of 17 group G sl reptococci isolated from symptomatic individuals between surveys had MICs greater than or equal to lllg/mL.
The prevalences of group G streptococci in the native Indian school population.mostly asymptomatic.for the three prevalence surveys were 7 .0. 5.9 and 4.8%.The prevalence from pharyngeal swabs obtained for the incidence survey from symptomatic school c hildren was 3.1 %. and for symptomatic adults and children below school age.5.0%.Thus there was no apparent association of group G streptococci with symptoms.suggesting that the CJythromycin -resistanl group G streptococci were unlikely to be contributing to pharyngeal disease in the community.
T-typing was performed on a subset of 59 of the group G streptococci.These included 29 susceptible strains -nine from the native Indian community and 20 from the Inuit -and 30 resistant strains.All 30 erythromycin-resistant strains were Tl6.Only two of 20 strains from the Inuit community could be typed with available antisera: one was T303 and one T305.Of the nine susceptible strains from the Indian community.four were T16.two 17. one each T305/307 and T7 /302.and one nontypeable.
Information documenting erythromycin usc in the two communities was not available.However.all antibiotics given in the two communities arc dispensed through the nursing station.Erythromycin ordered for lhe nursing station in the Inuit community was 120 100-tablct (250 mg) bottles.and 400 100-mL (40 mg/mL) bottles of suspension for 1985.F'or the native Indian community erythromycin acquisition was not available for Erythromycin-resistant group G streptococci 1985.but 150 100-tablet (250 mg) bottles.200 bottles of suspension and 10 inj ectables were orde red for 1987.

DISCUSSION
In the observation period a high proportion of isolates of group G streptococci resistant to erythromycin were documented in one of two geographically isolated northern communities surveyed.This high population prevalence of erythromycin resistance in group G streptococci has not.to the authors' knowledge.been previously reported.although erythromycin resistance is well described in some clinical isolates (7).E1ythromycin resistance was not observed in groups A or C streptococci.All resistant streptococci were of the same T-type.suggesting widespread dissemination of a single strain rather than a resistance determinant.This is cons istent wilh observations for group A slreplococci in which.when a high prevalence of erythromycin resistance has been observed.the res istant strains have been of the same serotype (1. 7.8).
While the high prevalence of erythromycin-res istant streptococci in the native Indian community is striking.no resistant group G streptococci were identified in the Inuit community.The Inuit community is as isolaLed as the native Indian community.and erythromycin use between the two communities as evidenced by antibiotic orders appeared sim ilar.During lhis same study year a s ignificantly higher prevalence and incidence of group A streptococcal disease was observed in the Inuit community.and the pharyngeal carriage of group G s treptococci was inversely proportional to that of group A streptococci in both communities (9).It is possible that lhe low prevalence of group A s treptococci in the native Indian community may have been one factor facilitating dissemination of the group G streptococcus.I lowever.a determination of what organism or community factors contributed to the emergence and widespread dissemination of this strain requires further investigation.
In a previous study of chemoprophylaxis for meningococcal meningitis in an isolated northern Inuit community the authors documented the rapid emergence and persistence of rifampinrcsistant I l aemoph ilus injluenzae following community-wide rifampin use (12).In this study a high prevalence of erythromycin-resistant group G streptococci with increasing prevalence of the strain throughout the study year was observed in one closed population.These studies suggest that patterns of anlimicrobial susceptibility in these communities may be highly community-specific.likely reflecting vruying patterns of antimicrobia l use as well as other factors not yet defined.These small.geographically isolated communities may serve as models for exploring parameters which promote the emergence and transmission of resistant organisms in non-institutionalized populations.
ACKNOWLEDGEMENTS: Dr Nicolle was a member of the Departments of Medicine and Microbiology and Infectious Diseases.University of Calgary when this study was perfom1cd.This study was funded by a grant from the National l leaJth Research and Development Fund. Hallie Subgut and Georgina Wood were community workers.Expert secretarial a ssistance was pro\1ded by B Kowalczyk.

Figure 1 )Fi. gure 2 )
Figure 1) Minimal inhibitory concentrations o.f groups A. C and G beta-hemolytic streptococci isolated in two northern Canadian communities to penicillin G. erythromycin and clindamycin