INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

The genus Corynebacterium presently comprises 46 species; 25 of them have been defined in the 1990s (8). It is the genus within the coryneform bacteria (i.e., aerobically growing, asporogenous, non-partially acid-fast, irregular, gram-positive rods) for which the largest number of species has been described so far. The description of new Corynebacterium species based on a polyphasic approach to taxonomy is the first step, but for clinical microbiologists, it is desirable that this is followed by the description of potential disease associations of the new taxa.

In the 1990s, the Department of Medical Microbiology, University of Zürich (DMMZ), has focused on the precise characterization of clinically relevant coryneform bacteria and their disease associations. During this process, 15 strains which were identified as Corynebacterium macginleyi were encountered. This species was defined in 1995 by Riegel et al. (12) as a result of their comprehensive investigations of lipophilic corynebacteria. Their study included three C. macginleyi strains, all of which were isolated from eye specimens. Since this original description not a single C. macginleyi strain has been described in the literature. Our present study outlines the clinical features of 15 patients from whom C. macginleyi was isolated, describes some additional characteristics useful in the identification of C. macginleyi, and for the first time, provides the antimicrobial susceptibility pattern of C. macginleyi. Most surprisingly, all 15 C. macginleyi strains were exclusively isolated from conjunctival swabs; we did not encounter C. macginleyi in any other clinical specimens, although clinically significant lipophilic corynebacteria from any kind of clinical material had been identified to the species level.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Culture conditions. The patients' material was cultured on Columbia agar plates (Becton Dickinson BBL, Cockeysville, Md.) supplemented with 5% sheep blood (SBA) and chocolate agar (Becton Dickinson) for 24 h at 37°C in a 5% CO₂-enriched atmosphere and on MacConkey agar (Becton Dickinson) at 37°C in ambient air. Fluid enrichment media were not used.

Biochemical identification. The lipophilic corynebacteria that were isolated were identified as outlined previously (8, 12, 15). The commercial API Coryne system (bioMérieux, Marcy l'Etoile, France), in conjunction with API Coryne database 2.0 (6), was used according to the manufacturer's instructions except that the strips were incubated for up to 48 h.

Susceptibility testing. Antimicrobial susceptibility patterns were determined by a microdilution method (Merlin Diagnostics, Bornheim-Hersel, Germany) as previously outlined in detail (2).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

The characteristics of the 15 patients from whom C. macginleyi was isolated are listed in Table 1. Between July 1992 and December 1997, 10 strains were isolated in DMMZ and 5 further isolates were referred to DMMZ; however, the majority of strains (13 of 15) were encountered in 1997; this could not be contributed to an increased awareness. All patients came from north and central Switzerland and were not epidemiologically linked. Records were available for 13 of the 15 patients. Seven patients were females and eight were males; the patients' mean age was 47 years (range, 3 months to 93 years). Six of the 15 patients had viral conjunctivitis with an obvious bacterial superinfection. Two of the patients had eyelid closure problems due to neurological disorders, two patients had foreign material (nasolacrimal duct catheter, contact lens) as predisposing factors, and one patient had a staphylococcal infection. For 4 of 15 patients, C. macginleyi was the only bacterial microorganism isolated, and for 3 of these 4 patients, the patients had bacterial superinfection of a viral conjunctivitis. For 7 of 15 patients, C. macginleyi was cultured together with coagulase-negative staphylococci (CoNS), for 3 patients it was cultured together with Staphylococcus aureus, and for 1 patient it was cultured together with alpha-hemolytic streptococci. Direct microscopy had been performed for only 3 of the 15 patients, and leukocytes as well as gram-positive rods were seen in all of them, although staphylococci grew also. Seven of the 15 patients were treated with quinolone-containing eye drops, and 6 other patients received different combinations of drugs including aminoglycosides, polymyxin B, bacitracin, and gramicidin. According to the physician in charge, the 3-month-old infant with purulent conjunctivitis improved only after treatment with oral amoxicillin-clavulanic acid. Chlamydia trachomatis was not detected with the ligase chain reaction kit (Abbott, Abbott Park, Ill.) in any of the patients tested. All patients recovered uneventfully after treatment.

The pathogenicity of C. macginleyi was difficult to determine, but we considered it to be rather low because we observed that the C. macginleyi strains from only four patients grew in pure culture. However, it seems that C. macginleyi may be able to cause bacterial superinfection on a compromised ocular surface.

The identification of lipophilic corynebacteria isolated from eye specimens is described in Table 2. C. macginleyi is one of the few corynebacteria not expressing pyrazinamidase activity (8, 12). Its ability to ferment mannitol is also not observed in many other corynebacteria (8, 12). When catalase-positive, small-colony-forming (<0.5 mm in diameter after 48 h of incubation) bacteria are found in eye specimens, one should always be suspicious that they are lipophilic corynebacteria. Lipophilism can be tested by comparing growth on SBA and SBA supplemented with 0.1 to 1.0% Tween 80 (Merck, Darmstadt, Germany), with lipophilic corynebacteria exhibiting colonies of up to 2 mm in diameter after 24 h of incubation on Tween-supplemented plates only. Propionibacteria, which might be isolated from the same clinical material, are not lipophilic. For the C. macginleyi strains included in this study we made two interesting observations: (i) 7 of 15 strains grew only very weakly on 1.0% Tween-supplemented SBA, whereas all 15 strains grew well on 0.1% Tween-SBA plates, and (ii) some strains exhibited a slightly rose pigment when they were grown on Tween-supplemented SBA. These two features might be of some use in the identification of C. macginleyi since they are not observed in any other lipophilic corynebacteria. The API Coryne system correctly identified all C. macginleyi strains (numerical codes: 1100105 [n = 2], 1100305 [n = 3], 1100315 [n = 1], 5100305 [n = 6], and 5100315 [n = 3]), whereas the species is not included in the present database of the commercial RapID CB Plus identification system (Remel, Atlanta, Ga.) (5).

The antimicrobial susceptibility pattern of C. macginleyi is given in Table 3. The selected antibiotics not only consisted of those used in the treatment of conjunctivitis but also consisted of other substances in order to provide data useful for the treatment of other C. macginleyi infections which might be observed in the future. As expected from the antimicrobial susceptibility patterns of other lipophilic corynebacteria, the MICs of -lactam antibiotics with the exception of that of ceftazidime were low for C. macginleyi. Quinolones, which had often been administered to our patients, also had low MICs. All strains were susceptible to tetracyclines, fusidic acid, rifampin, and glycopeptides. Two strains were resistant to macrolides, a feature also observed for the phenotypically closely related CDC group G bacteria (11, 13, 16). Antibiotics like aztreonam and fosfomycin were tested only because they may be used in the identification of corynebacteria or in media selective for corynebacteria. In summary, C. macginleyi is susceptible to all antimicrobial agents frequently used in the topical treatment of conjunctivitis. This may explain why C. macginleyi (or unspecified lipophilic corynebacteria) have not previously been reported in the literature as etiologic agents of bacterial conjunctivitis because the response to antimicrobial agents is, as in our patients, favorable. It may even be possible, although not recommended, to use antibiotics with high MICs because of the high concentrations which can be achieved by topical treatment.

In large-scale studies of the conjunctival flora of healthy persons, corynebacteria were always the second most frequently encountered bacteria after CoNS and were recovered from 11 to 58% of the persons: in the study of Høvding with 99 persons (ages, 15 to 39 years), CoNS were found in 62% of the persons and corynebacteria were found in 11% (9); Fahmy et al. (1) found CoNS in 82% of their test subjects and corynebacteria in 58%, but their test subjects were mainly elderly people (ages, >65 years) (1); Thiel and Schumacher (14) found that with increasing age there was a tendency toward a higher percentage of persons positive for corynebacteria. Perkins et al. (10) isolated corynebacteria from 12.1% of patients with conjunctivitis and from 7.3% of subjects in a healthy population. In all these studies except the one of Thiel and Schumacher (14), corynebacteria were never identified to the species level, which was probably due to the lack of availability of sufficient identification systems at that time. It is not unlikely that C. macginleyi may also have been part of the corynebacterial flora described in those studies.

It is striking that C. macginleyi had been isolated only from conjunctival swab specimens but not from any other clinical specimens, although all clinically significant lipophilic corynebacteria isolated from any clinical material had been identified to the species level by workers at DMMZ. As mentioned above, this is certainly not the result of an increased awareness but rather an indication of a certain tissue tropism which has also been described before for other coryneform bacteria, namely, Corynebacterium auris (ear samples) (4), Corynebacterium glucuronolyticum (genitourinary specimens from males) (3), and Turicella otitidis (ear samples) (7). The factors responsible for this tropism are not known at present.

In summary, up to now, only 18 C. macginleyi strains (including the ones in the present study) have been described in the literature. The incidence of C. macginleyi in conjunctival swabs from healthy persons is not known but will be the subject of a future prospective study by DMMZ. In the past, conjunctival flora had been studied in order to prevent perioperative eye infections through the use of appropriate antibiotics (14). Because of the antimicrobial susceptibility pattern of C. macginleyi, it can be predicted that this microorganism is covered by the currently used regimens.