Fluoroquinolone-Resistant Streptococcus pneumoniae in Spain: Activities of Garenoxacin against Clinical Isolates Including Strains with Altered Topoisomerases

Abstract

The activity of garenoxacin was assessed against 412 Streptococcus pneumoniae isolates (49.3% from the adult population). Overall penicillin, erythromycin, and ciprofloxacin (MIC, ≥4 μg/ml) resistance was 51.7, 35.4, and 1.5%, respectively. For all isolates, the garenoxacin MIC was ≤1 μg/ml. Amino acid replacements in GryA (Ser81→Phe or Tyr), ParC (Ser79→Phe or Tyr; Asp83→Gly; Lys137→Asn), and ParE (Ile460→Val; Asp435→Asn), alone or in combination, were ascribed to the reduced garenoxacin susceptibility (MIC range, 0.5 to 1 μg/ml) found in four isolates. The low impact of these mutations on garenoxacin activity envisages the possible coverage of S. pneumoniae populations resistant to preexisting quinolones.

Streptococcus pneumoniae plays an important role as etiological agent in respiratory tract infections, either community- or hospital-acquired. Moreover, antibiotic resistance in this pathogen is, in some cases, a matter of concern (22). In several countries, including Spain, penicillin and macrolide resistance has risen to nearly 50% and more than 30% of isolates, respectively, limiting the usefulness of these first-line antibiotic options, particularly in immunocompromised patients (2, 8). The emergence of fluoroquinolone-resistant pneumococcal strains may lead to another restriction in therapeutic choices (4, 20; R. Cantón, E. Loza, M. I. Morosini, J. Verhoef, and R. Jones, Abstr. 42^nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. E-61, 2002). Among clinically available fluoroquinolones, levofloxacin, moxifloxacin, and gatifloxacin are the most active against S. pneumoniae, although with variable in vitro activity and in vivo efficacy (5, 14).

Garenoxacin is a des-F(6) quinolone with an antibacterial spectrum that includes gram-positive bacteria (S. pneumoniae and the staphylococci), atypical respiratory pathogens (Legionella and Chlamydia species and Mycoplasma pneumoniae), and anaerobes (7, 21). Garenoxacin combines a high intrinsic activity with an enhanced pharmacodynamic profile and is also marginally affected by the PmrA-mediated efflux mechanism (3, 7). A balanced activity of this compound against DNA gyrase and topoisomerase IV has been observed in Staphylococcus aureus (11). Furthermore, garenoxacin appears to be less likely to select resistance, but if resistance mutants arise, they seem to retain a significant degree of susceptibility (9).

In this study, the antipneumococcal activity of garenoxacin was compared with that of ciprofloxacin, levofloxacin, moxifloxacin, penicillin G, cefotaxime, and erythromycin A in 412 prospectively collected clinical S. pneumoniae isolates (203 from the adult population) from several Spanish geographic areas. The correlation between susceptibility to quinolones and the presence of genetic determinants of macrolide resistance was assessed, and quinolone resistance mechanisms were characterized. S. pneumoniae isolates were collected between 1999 and 2001 in 14 Spanish hospitals. Overall clinical samples were distributed as follows: sputum (70 samples) and other lower respiratory tract secretions (59 samples), blood (50 samples), ocular secretion (61 samples), nasopharyngeal exudate (75 samples), middle ear fluid (83 samples), and other sources (14 samples).

Antibiotics were supplied as follows: garenoxacin (Bristol-Myers Squibb Pharmaceutical, Wallingford, Colo.); ciprofloxacin and moxifloxacin (Bayer Corporation, West Haven, Conn.); erythromycin A, cefotaxime, and levofloxacin (Aventis Pharma, Romainville, France); gatifloxacin (Grünenthal GmbH, Aachen, Germany); and penicillin G and reserpine (Sigma Chemical Co., St. Louis, Mo). MICs were determined by the agar dilution procedure using Mueller-Hinton agar (Oxoid Ltd., Basingstoke, United Kingdom) supplemented with 5% horse blood (Oxoid) and incubated overnight at 35°C in ambient air.

Four isolates for which garenoxacin MICs were outliers (0.5 to 1 μg/ml) were further characterized, and procedures to determine MICs, including that of gatifloxacin, were repeated by using the reference broth microdilution method (16). Eight other isolates (ciprofloxacin MIC range, 0.25 to 8 μg/ml) were tested as comparators. The presence of efflux activity against quinolones in these 12 isolates was phenotypically characterized by using the agar dilution procedure with and without a fixed concentration of 20 μg of reserpine/ml. These isolates were characterized by pulsed-field gel electrophoresis to discard possible clonal relatedness. Experimental conditions and pattern interpretation were as described previously (13, 19). The role of putative mutations present in the quinolone resistance-determining regions (QRDRs) of gyrase and topoisomerase IV coding genes of the four garenoxacin-outlier isolates and the eight comparator strains was assessed. PCR amplification of the QRDR regions of the cited genes was performed with the correspondent sets of primers (Sigma-Genosys Ltd., Cambridge, United Kingdom) (modified from the description in reference 17). PCR products were purified (QIAquick PCR purification kit, QIAGEN GmbH, Hilden, Germany) before being submitted for sequencing. The detection of erythromycin resistance genes was assessed in all isolates for which erythromycin MICs were ≥0.5 μg/ml, as reported (15).

The results of susceptibility testing are summarized in Table 1. Garenoxacin was the most active among the tested compounds, with an MIC at which 90% of isolates are inhibited of 0.03 μg/ml. Among tested pneumococci, penicillin and macrolide resistance were 51.7 and 35.4%, respectively, but this had no impact on the in vitro efficacy of garenoxacin. Garenoxacin MICs, plotted against the number of isolates, were clearly displaced to the left when compared with those of the other quinolones (Fig. 1). All isolates of pediatric origin were fully susceptible to all the quinolones tested (data not shown). In contrast, ciprofloxacin (3.0% resistance; MIC, ≥4 μg/ml) and levofloxacin (2.0% resistance; MIC, ≥8 μg/ml) resistance were found only among isolates from adult patients. Penicillin and erythromycin susceptibilities of these strains are shown in Table 2.

TABLE 1.

Comparative in vitro activities of garenoxacin against 412 Streptococcus pneumoniae isolates

Antibiotic	MIC (μg/ml)			% of fully susceptible isolatesa
	Range	50%	90%
Penicillin	≤0.008-4	0.12	2	48.3
Cefotaxime	≤0.004-2	0.03	0.5	91.3b
Erythromycin	≤0.01->64	0.03	>64	64.6
Ciprofloxacin	0.12-64	0.5	1	NDc
Levofloxacin	0.12-8	0.5	1	99.0
Moxifloxacin	0.01-2	0.12	0.12	99.0
Garenoxacin	≤0.008-1	0.03	0.03	ND

^aNCCLS breakpoints (document M100-S12).

^bNCCLS meningitis criterion.

^cND, not defined. For ciprofloxacin, a susceptibility breakpoint of ≤2 μg/ml was used (98.5% susceptibility). For garenoxacin, 100% of isolates were inhibited by 1 μg of this compound per ml.

FIG. 1.

MIC distribution of garenoxacin and comparator quinolones against 412 S. pneumoniae isolates.

TABLE 2.

Characteristics and changes in the QRDR of gyrase and topoisomerase IV of isolates with reduced susceptibility to garenoxacin and comparator strains

Isolate no.	Specimen type	Serotype	Patient type	MIC (μg/ml)a							Amino acid change(s)
				PEN	ERY	CIP	LEV	GAT	MOX	GAR	gyrA	parC	gryB	parE
1	Blood	15 A	Adult	0.25	>64	64	32	8	4	1	Ser81→Phe	Ser79→Phe		Ile460→Val
2	Sputum	23F	Adult	4	0.03	64	32	8	4	0.5	Ser81→Tyr	Ser79→Tyr, Lys137→Asn		Ile460→Val
3	Pleural	38	Adult	0.01	0.12	32	16	4	4	0.5	Ser81→Phe	Ser79→Tyr
4	Sputum	19	Adult	0.5	1	16	16	8	2	0.5	Ser81→Phe			Asp435→Asn
5	Sputum	6B	Adult	2	>64	8	2	0.5	0.25	0.06		Ser79→Phe
6	Sputum	6B	Adult	2	64	4	2	0.25	0.12	0.03		Ser79→Phe
7	BALb	23F	Adult	2	0.03	2	1	0.25	0.12	0.03		Asp83→Gly, Lys137→Asn		Ile460→Val
8	Blood	19	Pediatric	0.06	1	1	1	0.25	0.12	0.03		Lys137→Asn		Ile460→Val
9	BALb	11	Pediatric	0.01	0.03	1	1	0.5	0.25	0.06				Ile460→Val
10	Pleural	11	Adult	0.01	0.03	1	1	0.25	0.25	0.06				Ile460→Val
11	Otic	14	Adult	2	>64	0.5	0.5	0.12	0.06	0.03
12	Otic	1	Pediatric	≤0.008	≤0.01	0.25	0.5	0.12	0.12	0.03

^aNCCLS broth microdilution method. PEN, penicillin; ERY, erythromycin; CIP, ciprofloxacin; LEV, levofloxacin; GAT, gatifloxacin; MOX, moxifloxacin; GAR, garenoxacin.

^bBAL, bronchoalveolar lavage.

Overall, erm(B) was the prevalent resistance gene (29.8%) among erythromycin-intermediate and -resistant isolates. Among isolates from adult patients harboring erm(B) as the sole resistance determinant (36%), 2.0% were resistant to ciprofloxacin and 1.0% were resistant to levofloxacin. The mef(A) gene alone was found in only 3.6% of the whole studied population, but no resistance to any quinolone tested was observed within this group. The simultaneous presence of both genes was almost negligible (1% of the total population).

The characteristics of the 12 selected isolates and the correspondent QRDR analysis are listed in Table 2. The pulsed-field gel electrophoresis analysis revealed that the 12 isolates were clonally unrelated (not shown). Despite the reduced garenoxacin susceptibility of the aforementioned four isolates (MIC range of 0.5 to 1 μg/ml), this was the most active quinolone tested, followed, in order of activity, by moxifloxacin (2 to 4 μg/ml), gatifloxacin (4 to 8 μg/ml), levofloxacin (16 to 32 μg/ml), and ciprofloxacin (16 to 64 μg/ml). The inhibitory activity of the tested quinolones was not significantly modified, at least phenotypically, with the addition of reserpine (data not shown). The increase in garenoxacin MICs for these four isolates could be accounted for by the observed changes within the QRDRs of the targeted-protein subunits, while the impact of active efflux appears to be marginal. When QRDRs were analyzed, either Phe or Tyr replaced Ser at position 81 of the GyrA subunit. When this change coexisted with the Ser79 replacement in ParC (also by Phe or Tyr), quinolone MICs rose to maximum values, particularly those of ciprofloxacin (64 μg/ml) and levofloxacin (32 μg/ml) (Table 2). The concomitant presence of Lys137→Asn in ParC (isolate 2) did not seem to add an increase in phenotypic resistance. In the case of ParE, two isolates (number 1 and 2) exhibited the frequent Ile460→Val replacement, considered to be neutral (6). Isolate number 4, for which the MIC of garenoxacin increased by nearly 20-fold (0.5 μg/ml, compared with the modal MIC of 0.03 μg/ml of the collection), had the amino acid change Asp435→Asn combined with Ser81→Phe in GyrA. Isolate 3 had no changes in the ParE subunit, and no isolate exhibited alterations in the GyrB subunit. None of these combinations of amino acid replacements was responsible for garenoxacin MICs higher than 1 μg/ml. The presence of mutations outside the QRDRs and their potential contribution to resistance cannot be ruled out in these strains.

None of the comparator isolates have amino acid replacements in either GyrA or GyrB subunits. The phenotypic transcendence of changes in ParC and/or ParE subunits (Table 2) of these strains has been extensively documented elsewhere (6, 12). Isolates for which ciprofloxacin MICs were 0.5 and 0.25 μg/ml, (isolates 11 and 12, respectively) had no mutations in the regions of the four genes analyzed.

The increasing prevalence of quinolone resistance among S. pneumoniae in certain countries, including Spain, should lead to the consideration of a reduction in empirical therapeutic options for the treatment of infectious processes, mainly in those where a severe course is observed (5, 18). With the aim of evaluating the potentiality of garenoxacin for the treatment of pneumococcal infections, a nationwide surveillance study was designed, and 412 S. pneumoniae isolates, prospectively collected in different Spanish hospitals, were tested. Garenoxacin was the most active of the antibiotics assayed, and no resistance to this compound was observed, while penicillin, erythromycin, and ciprofloxacin resistance was found in 51.7, 35.4, and 1.5%, respectively, of the entire studied population.

Unlike in other countries where resistant clones are responsible for quinolone resistance dissemination (10), in Spain, both nonclonal dispersion and preponderance of certain quinolone-resistant pneumococcal clones have been reported (1). The absence of clonal relatedness among the 12 isolates fully characterized in the present study suggests that despite the small sample size, quinolone resistance appears to be of nonclonal origin in the studied population.

Huge antibiotic consumption is, at least in part, deemed to be responsible for the rise in antibiotic resistance (8). Even more strikingly, the use of unrelated antibiotics may lead to the coselection of diverse resistance traits. This may partially explain the tendency observed in the quinolone-resistant isolates of the population studied to cluster among macrolide- and β-lactam-nonsusceptible isolates from adult patients (C. García-Rey, L. Aguilar, F. Baquero, and the Spanish Surveillance Group for Respiratory Pathogens, Letter, Antimicrob. Agents Chemother. 44:3481-3482, 2000). The suitable choice and dosage of antibiotic compounds contribute to curtail overrepresentation and subsequent dispersion of successful, frequently multiresistant, pneumococcal clones. The extended use of ciprofloxacin and levofloxacin may have contributed to the expansion of resistant clones harboring second-step mutations added to preexisting ParC alterations. Garenoxacin has not yet been exposed to the clinical environment. Considering the balanced interaction of this compound with both gyrase and topoisomerase IV (11), it is of interest to witness the impact of the presently described and hitherto-unknown mutations on garenoxacin efficacy and to determine if its use could minimize the prevalence of these strains harboring such mutations, thus ameliorating the overall performance of earlier quinolones. Clinical evidence is still required to validate this approach.