The properties of erythromycin resistant staphylococcal strains obtained from clinical material and those so created in the laboratory are quite different. Thus, whilst clinical strains resistant to erythromycin have sometimes been obtained in large numbers after the lavish use of erythromycin (e.g. Forfar et al., 1966), the explanation for this cannot be based on laboratory experiments such as those described here. We must, therefore, consider other mechanisms. Resistance to most antibiotics, both in Staph, aureus and in Enterobacteriaceae, is thought to have arisen by the freak occurrence of resistant strains that contain genetic information enabling the cell to resist specific antibiotics. With the use of such antibiotics, there has been selection of these resistant isolates at the expense of sensitive organisms. There has also been a certain amount of ‘infection’ of sensitive bacteria with those genes (see Lacey, 1975a, b J. Such an explanation probably applies to erythromycin resistance in Staph, aureus, where a few (or even one) strains that contained the complex mechanism for resistance to erythromycin, have been selected with the use of the drug. We believe that clinical strains of staphylococci have evolved rapidly in nature, and this is shown by alterations in phage typing pattern, and in gain or loss of a variety of antibiotic resistances (Jevons, John and Parker, 1966; Lacey, 1975a). The isolation of a number of erythromycin resistant strains with rather variable properties, is certainly consistent with the rarity with which this resistance mechanism is thought to occur in nature. The important inference of these considerations is that the frequent isolation of pathogens resistant to erythromycin that occurred in the 1950's and 1960's was due to the repetitive collection of essentially one or a few resistant strains, and was not due to the appearance of erythromycin resistance arising de novo in many strains. This concept applies to other resistances in Staphylococcus aureus with the intensive use of almost any antibiotic (e.g. tetracycline, ampicillin, fusidic acid, neomycin or gentamicin), that is followed by widespread resistance. For all these resistances, the mechanism is probably as for erythromycin. How relevant are these considerations for Haemophilus influenzae, Streptococcus pneumoniae and Streptococcus pyogenes? The resistance to the antibiotic in these organisms, including that to erythromycin, has always been rare (the author has never seen an erythromycin resistant Group A streptococcus, and these are routinely tested for sensitivity in his laboratory). The rarity of resistance in these organisms is probably due to the absence of naturally occurring resistance genes in each of the species. As with Staphylococci, it has been impossible to create in vitro strains highly resistant to erythromycin de novo, and any slightly resistant organisms have been defective (and hence probably non-pathogenic). The risk that these pathogens will acquire resistance during therapy now seems remote and even if any of these did, they would: (a) probably be non-pathogenic to the host in question, and (b) not become an epidemic problem, because they would be at a grave disadvantage compared with sensitive bacteria. (Similarly, the occasional reports of resistant bacteria appearing after prolonged use of erythromycin have rarely indicated whether such strains are pathogenic.) Thus, in summary, the use of erythromycin should not be governed by the fear that this drug is particularly prone to select resistance. In fact, the reverse is probably true as it is not involved in the potentially most dangerous of all resistances-the transfer of resistance genes from a commensal E. coli, etc., to dangerous pathogens, such as Salmonella typhi This transfer is encouraged by broad spectrum antibiotics, e.g. tetracyclines and ampicillin when the gut coliforms are often inadvertently exposed when these antibiotics are used (often inappropriately!) for respiratory, soft tissue or urinary tract infections.
Microbiological grounds for antimicrobial treatment of experimental pseudomonal keratitis
