The t h r e a t o f t h e e m e r g e n c e of antimicrobial-resistant Gram-positive pathogens in C a n a d a

DE The threat of the emergence of antimicrobial-resistant Gram-positive pathogens in Canada. Can J Infect Dis 1994;5(Suppl C):9C-14C. Since lhc early 1980s. much attention has been focused on lhc emergence or resistance in nosocomially acquired Gram-negative pathogens. However. in lhc 1990s we are witnessing in North America lhc development and spread or mult iple resistance in Gram-positive pathogens in the hospital selling as well as in lhc community. Mclhicillin rcsislanl Staphylococcus aureus and vancomycin-resislanl enlerococci are now endemic in many urban centres in lhc Uni led Stales. although less so in Canada. In some states. penicillin-rcsislanl Streptococcus pneumoniae in lhc community selling has gone from rates of less than 5% in 1988 to 50% in 1994. including: resistance lo lhircl-generalion cephalosporins and carbapcncms. Although these same pathogens have now been identified in Canada. we may still be in a position lo limit or prevent their spread.

lowed rapidly by the identification of penicillinaseproducing staphylococci. After just 10 years of penicillin use. 73% of Staphylococcus aureus isolates from in-patients at the Boston City Hospital were penicillin-resistant (1). This is one of the earliest illustrations that one of the consequences of the introduction of any new antimicrobial agent is the development of bacterial resistance to its action (2.3). Such resistance may arise by a mutation that reduces target affinity or allows the overproduction of a drug modifying enzyme. However, U1e introduction into a bacterium of foreign DNA encoding for resistance may bypass the need for endogenous mutational events. This DNA may be introduced into the chromosome by transfom1ation and recombination or may be integrated into the bacterial cell on plasmids. which may be transferable from one organism lo another by conjugation, transduction or transformation (4 .5). In this article we review the emergence of methicillin-resislanl S aureus (MRSA). multiply resistant Enterococcus species and penicillin-resistant Streptococcus pneumoniae (PRSP). and discuss the threat they may pose in Canada.

METHICILLIN -RESISTANT STAPHYLOCOCCUS AUREUS
Soon after methicillin became available, resistance lo it was reported in S aureus. This resistance was not the result of destruction of the antibiotic by a betalactamase (6). IL was subsequently found that the resistant strains had a newly acquired resistance gene. mecA, that encoded for a penicillin binding protein (PBP). PBP-2a (7.8). PBP-2a is able to maintain cell wall integrity during growth and division when native enzymes (PBPs) needed for assembly of the cell wall are inactivated by beta-lactam antibiotics. The spread of staphylococcal clones carrying the mecA gene has resu I ted in the worldwide dissemination of MRSA (9). Outbreaks reported in the United States in the 1970s were confined primarily to large, tertiary care teaching hospitals (10,11) . However. in the 1980s some community hospitals and rehabilitation and extended care facilities experienced an increasing prevalence of MRSA colonization and/or infection (l l -13). For instance, the prevalence of MRSA increased dramatically in Veterans· Administration medical centres between 1975 and 1984 (14). MRSA continues lo be endemic in many of these facilities.
In Canada, the prevalence of MRSA remains low relative to other countries, but varies in different provinces, cities and hospitals. A point prevalence survey to determine MRSA colonization rates at three university-affiliated tertiary care facilities in downtown Toronto found that none of 1219 patients were colonized with MRSA (personal communication). McArthur et al (15) surveyed 20% of residents in each of 132 long term care facilities in Ontario for colonization with MRSA. One or more MRSA colonized residents were identified in eight of 132 facili-lOC Y·DONOTCGn ties (6. 1 %, 95% confidence level 1.9 to 10.2%). Only two facilities had previously recognized the presence of MRSA. Taylor et al (16) at the University of Alberta Hospitals found that the majority of MRSA isolates from hospitalized patients was community acquired and that most were isolated from residents of one aboriginal community. A study done by Embil et al (17) reviewed the experience with five tertiary care leaching hospitals on the Canadian prai!ies, and found similar results. Patients usually had MRSA identified at admission: in only one of five centres was the majority of isolates acquired nosocomially. Patients with MRSA present al admission were more frequently aboriginal. It appears that in Canada the major reservoirs for MRSA may be outside the hospital setting.
Can MRSA be controlled once introduced into a hospital? Boyce (18) reviewed 46 published outbreaks and found that early implementation of control measures was associated with success in eradication. All 11 hospitals with 20 or fewer cases were successful in eradicating MRSA compared with only 71% with 20 to 39 cases and 10% with 40 or more cases. In Sarnia, Ontario 21 patients with MRSA were identified in a secondary care hospital between March 1990 and January 1991 (19). The reservoir was suspected to be in nursing homes in the community. A survey of all hospital patients and nursing home residents in Samia was carried out, and those found to be colonized with MRSA were treated. All subsequent admissions from other institutions were screened for MRSA and treated if colonized. Since the institution of these policies no further outbreaks have occurred as of January 1994 (personal communication). Policies requiring admission screening of patients at risk for MRSA colonization (eg. admitted from an area where MRSA is endemic) may control the spread of this organism in Canada.

MULTIPLY RESISTANT ENTEROCOCCUS SPECIES
The enterococcus is an important nosocomial pathogen despite its low virulence: in National Nosocomial Infections Surveillance System (NNIS) hospitals. enterococci are the third most common pathogen associated with nosocomial bloodstream infections and the second most commonly isolated nosocomial pathogen overall (20-23). This is due to llie organism's ubiquitous nature, its inherent antimicrobial resistance. and its ability to acquire multiple resistance traits.
Acquired high-level resistance lo the aminoglycosides was first reported in the 1970s. By the mid -1980s many reports had documented high -level aminoglycoside resistance in both Enterococcusjaecalis and Enterococcus Jaecium (24-26). Resistance to betalactams has also dramatically increased during the past decade (27-30). This resistance is usually a result of an increase in U1e amount of low affinity PBPs syntl1esized by the cell (31). However, high-level ampicillin resistance may also occur due lo the acquisition of
Studies of lhe genetics and mechanisms of glycopepticle resistance in enterococci have suggested three classes of resistance: A. B and C. All classes have genes lhal encode for a cell wall precursor wilh a reduced affinily for glycopcptide antibiotics (43) . Isolates wilh resistance lo high levels of vancomycin (mimimum inhibitory concentration [MIC] 256 pg/mL or greater) and lo leicoplanin (MIC 16 pg/mL or greater) have been classified as phenolypic class A. This is lhe mosl prevalent phenotype. ll is encoded for by lhe vanA gene. which may be lransleITed between strains on plasmids and lransposons. Class B strains are resislanl lo vancomycin. with MICs ranging from 16 lo 1024 pg/mL. bul are susceptible lo leicoplanin (MI C less t.han 8.0 pg/mL). This type of resistance is encoded for by lhe vanB gene. which is also transferable. Cla s C resistance. found in Ente rococcus gallinarium and Enterococcus casseliflavus. comprises constitutive low-level resistance lo vancomycin (MIC 4.0 pg/mL or greater, or 32 ~tg/mL or less) and susceptibility lo leicoplanin (43)(44)(45). This lype of resistance is chromosomal. intrinsic and nontransferable. The study of lhe rapid spread of glycopcplideresislanl enlerocoeei in Europe and lhe United Stales suggests Lhal dissemination occurs nol only by clonal spread of individual strains, but. also by lhe horizontal spread of resistance delem1inanls from strain lo strain (36,40,46). The rapid worldwide emergence ofvancomycin resistance in enlerococci more than 35 years after the inlroduelion of lhis antibiotic probably ref1ecls selective pressure clue lo ils increased use (47,48). For instance, Ena el al (48) found thal the rale ofvancomycin use al a university hospital increased from 5. 7 g/ I 000 patient days in I 981 lo 21 g/ 1000 palienl days in 1991.
In Canada. bolh high -level aminoglycoside-resist.anl and bela-laclamase-ncgalive ampicillin resistant enlerococci have been reported (49 .50). There has been only one case report of a bela-laclamase-producing EJaecalis isolate (51). The firsl vancomycin-resislanl isolate of E Jaecium was described by Kibsey el al (52). This organism was isolated from lwo patie nts in an intensive care unil. Both patients were immunocompromised and had received broad spectrum antimicrobial therapy.
The emergence and dissemination of multiply resislanl enlerococci is resulting in infections for which there are no unifom1ly effeclive lherapeu lie options.
The possibility that lhc vanA gene will lransfer inlo other Gram -positive organisms such as MRSA and/or PRSP is of even great.er concern. The transfer of vancomycin resistance from enterococci to S aure us has already been accomplished in vivo in lhe laboratory (53). VRE are readily delectable in the laboralory if appropriate antimicrobial testing practices are employed (54.55). Once identified , control of lhe spread or this organism will require strict adherence lo infection control practices and the institution of a surveillance program lo identify colonized patients and health care workers. Othe rwise whal has happened in lhe United Stales. and in particular in New York City, could easily occur in Canada (36).

PENICILLIN -RESISTANT STREPTOCOCCUS PNEUMONIAE
Of equal concern is the rapid emergence of l'f~SP in North America. In 1967. lhc first patient infected wilh a PRSP strain wilh inlermediale resislancc lo penicillin (MIC 0.1 to 1.0 pg/mL) was reported from Australia (56). Subsequently. other strains were identified in New Guinea and Au lralia (57.58). In 1977 in South Africa. outbreaks of the firsl strains lhal were highly pcnicillin -resislanl (MIC 2 pg/mL or greater) were reporled (59) , as was the first multiply resislanl strain (resislanl lo penicillin. tetracycline, erylhromycin. clindamycin. lrimethoprim-sulphamelhoxazole and chloramphenicol) (59.60). PI~SP infection in the United Stales was initially reported in 1974 (61). and the first resisla nl Canadian isolates were d escribed in the mid -l 970s (62). Before 1987. in lhe United Slat.cs only 5% of pneumococci demonstrated intermediate resista nce and less lhan 1 % we re highly resislanl (63). The last two Canadian surveillance studies. pcrfom1ed in th e 1980s. found less than 1.5% of isolates lo have intermediate rcsislance and none to be highly resistant (64.65). Rcccnlly. however. disturbingly high rates of PHSP (more lhan 50%) have been identified in Spain and Hungary (66.67) . and several reports from the United Stales document increases in the number of PHSP. From 1987 lo 1992. the proportion of S pne umoniae strains submit.led to lhe Cenlers for Disease Conlrol and Prevention lhal were highly resistant to penicillin increased from 0.02% Lo 1.3% (68). In community surveys in Kentucky and Tennessee in 1993 il was found thal 65% and 39%, respectively. of S pneumoniae strains were PRSP. and of these 65% and 30%. respectively. were highly rcsislanl lo penicillin (69).
Resistance lo penicillin in clinical isolates of S pne umoniae is clue lo the development of PBPs lhal have greatly decreased affinity for U1e anlib iot.ic (70)(71)(72). In those isolates that. have the highest levels of resistance lo penicillin. there have been reductions in lhe affinity of al least four of the five h igh molecular weight PBl'S (73). Low alTinity forms are believed lo ha ve arisen by localized interspecies recombinationa l events that re-llC Low eta/ USE place parts of the PBP gene with the corresponding parts of the homologous genes of closely related species (74). That is, the S pneumoniae has acquired PBP genes from other streptococci that are resistant to penicillin. Thus, the evolution of resistance in S pneumoniae is somewhat different from that in MRSA or enlerococci. The mecA and vanA genes are highly conserved, and their spread occurs either by clonal dissemination of a strain carrying them (as usually occurs with MRSA). or by U1eir transfer from strain to strain on a plasmid or transposon (as occurred with enterococci in New York City) (75,76). Although clonal dissemination of strains of PRSP has occurred (77), the transformation and recombination events that transfer penicillin resistance to S pneumoniae are not uncommon. and the emergence of PRSP may be due lo the repeated new development of low affinity PBPs (78).
In Canada there has been a recent increase in reports of PRSP. In an unselected sample of 2642 isolates of S pneumoniae collected from across Canada, 252 (9.5%) were found lo be resistant to penicillin by the oxacillin screen test. Of the resistant isolates, 39 were from blood cultures. four from sterile fluids and 209 from the respiratory tract. One hundred and twentyseven of these were from Ontario, 31 from British Columbia. 34   12C Y·DONOTCOPY (seven) and 6B (six) (personal communication). The appearance of PRSP in Canada has several important implications. First, if rates of resistance in a community become greater than 3%, then first-line low-cost antimicrobials for empirical therapy (ie. penicillin) may have to be replaced with more costly agents. Second. beta-laclams other ilian penicillin have reduced activity against. PRSP isolates, thereby potentially limiting the use of a whole class of antimicrobials (79,80). Finally. alternative antimicrobials to penicillin for the treatment of invasive infections due lo high-level resistant S pneumoniae are not as effective as penicillin for p nicillin -suscepUble strains, so that. ilierapy for patients with these infections may be compromised (81.82). Effective alternative antimicrobials for invasive disease due lo PRSP in nonmeningitis cases are ceftriaxone and cefotaxime.

C ONCLUSIONS
There are several strategies to try lo combat the problem of emerging antimicrobial resistance. One approach is lo reduce selective pressure for the development of resistance by more prudent use of antimicrobials. Equally important. is the need for belt.er surveillance to determine the frequency of antimicrobial resistance so iliat studies can be designed lo determine the import.ant fact.ors in the emergence, persistence and transmission of drug-resistant organisms. Ongoing transmission of organisms being introduced may then be cont.rolled. As Cohen (3) noted. "unless currently effective antimicrobial agents can be successfully preserved and the transmission of drug-resistant organisms curtailed, the post-antimicrobial era may be rapidly approaching".