In Vitro Capability of Faropenem To Select for Resistant Mutants of Streptococcus pneumoniae and Haemophilus influenzae

Abstract

When tested against nine strains of pneumococci and six of Haemophilus influenzae of various resistotypes, faropenem failed to select for resistant mutants after 50 days of consecutive subculture in subinhibitory concentrations. Faropenem also yielded low rates of spontaneous mutations against all organisms of both species. By comparison, resistant clones were obtained with macrolides, ketolides, and quinolones.

Streptococcus pneumoniae is the leading bacterial cause of community-acquired pneumonia. The incidence of pneumococci resistant to penicillin G and other β-lactam and non-β-lactam compounds is increasing worldwide (8, 10). Although introduction of a pediatric conjugated vaccine has led to a significant decrease in systemic infections caused by drug-resistant pneumococci, this decrease has not been mirrored in pneumococcal community-acquired respiratory tract infections caused by resistant strains. Also, nonvaccine serotypes with raised penicillin G MICs have recently appeared (15). There is thus still a need for new agents to treat these infections.

Haemophilus influenzae is a second major cause of community-acquired respiratory infections in children and adults. In countries such as the United States, where the H. influenzae type b vaccine is widely used, H. influenzae type b has been replaced by untypeable H. influenzae strains (18). The major resistance mechanism in H. influenzae in the United States and Europe is β-lactamase production (TEM-1 and ROB-1). The incidence of β-lactamase-negative, ampicillin-resistant (BLNAR) strains in the US is <1% (18). The incidences of BLNAR strains are high in France and Japan and are reportedly on the rise in other countries (18). Of the β-lactams available for treatment of infections caused by this organism, cefixime and cefpodoxime are the most potent, followed by amoxicillin-clavulanate and cefuroxime (18). Despite relatively low MICs, the pharmacokinetic and pharmacodynamic properties of macrolides and ketolides cast doubt on their clinical efficacy against H. influenzae (18).

Faropenem medoxomil (6, 9, 16, 17) is an oral penem to be introduced for oral treatment of pediatric and adult community-acquired respiratory tract infections. This compound has low MICs against S. pneumoniae as well as H. influenzae (16, 17). This study uses single- and multistep methodologies to examine the capability of faropenem compared with those of amoxicillin-clavulanate, cefuroxime, cefdinir, azithromycin, telithromycin, levofloxacin, and moxifloxacin to select for resistant mutants of nine S. pneumoniae and six H. influenzae strains of various resistotypes. Intravenous carbapenems, such as imipenem and meropenem, were not included, because we felt that the drugs tested should all be orally available.

For testing of S. pneumoniae strains, three penicillin-susceptible, three penicillin-intermediate-resistant, and three penicillin-resistant strains were chosen. Of these, three were erythromycin susceptible; one strain had erm(B), two had mef(A), one had erm(B) plus mef(A), and one had an L4 and one a 23S rRNA mutation. Three of the nine strains were quinolone resistant with defined mutations in their quinolone resistance-determining regions. None of the three quinolone-resistant pneumococci exhibited quinolone efflux according to the reserpine method.

Among the six H. influenzae strains, three were β-lactamase negative, two were β-lactamase positive, and one was BLNAR with defined mutations in PBP3. Among these six strains, one organism was macrolide hypersusceptible through lack of an efflux mechanism and the remaining five were efflux pump positive but macrolide susceptible by current CLSI criteria. Strains were frozen at −70°C in skim milk (Difco Laboratories, Detroit, MI) before use. Faropenem powder (the sodium salt) was obtained from Replidyne, Inc., Louisville, CO, and other drugs were obtained from their respective manufacturers.

Multistep resistance selection techniques were as described previously by our group (1-4, 7, 12-14), using Mueller-Hinton broth plus 5% lysed horse blood (5) for pneumococci and freshly made Haemophilus test medium for H. influenzae. Daily serial passages from each strain were performed with subinhibitory concentrations of all antimicrobials. For each subsequent daily passage, an inoculum was taken from the tube at 1 dilution below the MIC. The above-mentioned inoculum was used to determine the next MIC. Daily passages were continued until a significant (>4-fold) increase in MIC was found. The drug concentrations tested are listed in Tables 1 and 2. A minimum of 14 passages was performed in every case. The maximal number of daily passages was 50. The stability of the acquired resistance was determined by MIC determination after 10 daily passages of selected clones on drug-free agar. The identities of parents and resistant clones were confirmed throughout the study by pulsed-field gel electrophoresis (1-4, 7, 12-14).

TABLE 1.

Faropenem S. pneumoniae multistep selection results^a

Strain	Antibiotic	Initial MIC (μg/ml)	Result for selected resistance		Retest MIC after 10 antibiotic-free subcultures								Resistance mechanism mutation(s)
			MIC	No. of passages	Faropen	A-C	Cefur	Cefdin	Azith	Teli	Levo	Moxi
3009	Faropen	0.03	0.06	50
	A-C	0.016	0.03	50
	Cefur	0.03	0.125	50
	Cefdin	0.125	0.06	50
	Azith	>64	NT	NT
	Teli	0.016	0.03	50
	Levo	1	16	18	0.03	0.008	0.06	0.06	>64	0.016	32	4	GyrA(S81Y); ParC(S79Y)
	Moxi	0.25	2	19	0.03	0.03	0.06	0.125	>64	0.008	2	4	GyrA(S81F); ParC(D83N)
1076	Faropen	0.03	0.06	50
	A-C	0.016	0.03	50
	Cefur	0.03	0.125	50
	Cefdin	0.125	0.125	50
	Azith	>64	NT	NT
	Teli	0.5	>64	14	0.016	0.016	0.03	0.125	>64	>64	8	1	Not found
	Levo	4	16	50
	Moxi	1	8	35	0.03	0.03	0.06	0.06	>64	1	32	4	ParC(S79Y)
1077	Faropen	0.03	0.03	50
	A-C	0.016	0.03	50
	Cefur	0.03	0.125	50
	Cefdin	0.125	0.5	50
	Azith	0.03	>64	29	0.016	0.03	0.03	0.06	>64	0.125	32	4	23S rRNA(A2058G)
	Teli	0.004	0.25	15	0.016	0.016	0.03	0.125	>64	0.25	16	4	23S rRNA(A2058G,
	Levo	16	32	50									A2106G)
	Moxi	4	4	50
3665	Faropen	0.125	0.25	50
	A-C	0.125	0.5	50
	Cefur	0.125	0.5	50
	Cefdin	0.125	0.5	50
	Azith	8	16	50
	Teli	0.125	2	14	0.125	0.06	0.25	0.25	4	0.25	0.5	0.125	Not found
	Levo	0.5	2	50
	Moxi	0.125	1	38	0.125	0.25	0.25	0.25	8	0.125	4	1	GyrA(S81Y)
3676	Faropen	0.125	0.125	50
	A-C	0.03	0.125	50
	Cefur	0.125	0.25	50
	Cefdin	0.03	0.06	50
	Azith	4	8	50
	Teli	0.06	2	14	0.125	0.03	0.25	0.06	2	0.06	1	0.5	Not found
	Levo	0.5	2	50
	Moxi	0.25	8	50	0.125	0.06	0.25	0.06	4	0.06	16	8	GyrA(E85G); GryB(S478I); ParC(S79Y)
3243	Faropen	0.125	0.125	50
	A-C	0.06	0.25	50
	Cefur	0.25	1	50
	Cefdin	0.25	1	50
	Azith	0.03	0.25	33	0.125	0.125	0.5	0.5	0.5	0.06	0.5	0.25	L22(INS 93SFRPRA94)
	Teli	0.004	0.06	29	0.125	0.125	0.5	0.5	>64	0.06	1	0.25	Not found
	Levo	0.5	2	50
	Moxi	0.125	1	20	0.125	0.06	0.5	0.5	0.016	0.016	1	1	Not found
24	Faropen	1	2	50
	A-C	1	2	50
	Cefur	2	4	50
	Cefdin	2	8	50
	Azith	>64	NT	NT
	Teli	0.25	32	14	1	2	4	4	>64	16	1	0.125	L4(G69R)
	Levo	1	2	50
	Moxi	0.25	4	50	1	1	4	4	>64	0.25	32	4	GyrA(S81Y); GyrB(P413S); ParC(S79F)
3413	Faropen	1	1	50
	A-C	2	8	50
	Cefur	4	8	50
	Cefdin	2	8	50
	Azith	>64	NT	NT
	Teli	0.03	0.25	30	1	4	8	4	>64	0.5	0.5	0.25	Not found
	Levo	0.5	1	50
	Moxi	0.125	0.25	50
2527	Faropen	1	1	50
	A-C	1	2	50
	Cefur	8	32	50
	Cefdin	8	16	50
	Azith	0.016	4	20	1	2	8	16	4	0.03	32	4	Not found
	Teli	0.004	0.016	50
	Levo	16	32	50
	Moxi	2	8	50

^aCross-reactivity is denoted in bold. Abbreviations: Faropen, faropenem; A-C, amoxicillin-clavulanate; Cefur, cefuroxime; Cefdin, cefdinir; Azith, azithromycin; Teli, telithromycin; Levo, levofloxacin; Moxi, moxifloxacin; NT, not tested; INS, insertion.

TABLE 2.

Faropenem H. influenzae multistep selection results^a

Strain	Antibiotic	Initial MIC (μg/ml)	Result for selected resistance		Retest MIC after 10 antibiotic-free subcultures								Resistance mechanism mutation(s)
			MIC	No. of passages	Faropen	A-C	Cefur	Cefdin	Azith	Teli	Levo	Moxi
110-019	Faropen	2	4	50
	A-C	2	2	50
	Cefur	2	4	50
	Cefdin	1	1	50
	Azith	1	8	15	2	2	2	1	8	8	0.03	0.06	Not found
	Teli	1	8	21	2	2	2	1	2	4	0.03	0.06	L4(Q61R)
	Levo	0.03	0.25	28	1	2	2	0.5	1	2	0.125	0.5	GyrA(D88N)
	Moxi	0.06	0.5	39	1	2	2	0.5	1	2	0.125	0.25	GyrB(E469D)
HI-30	Faropen	1	2	50
	A-C	2	2	50
	Cefur	2	4	50
	Cefdin	0.5	0.5	50
	Azith	0.25	4	20	2	2	2	1	4	2	0.03	0.06	Not found
	Teli	0.5	4	24	2	2	2	0.5	1	4	0.03	0.03	L4(Q61K), L22(K90E)
	Levo	0.03	0.5	28	2	2	2	0.5	0.25	0.5	1	1	GyrA(S84L), ParE(S458L)
	Moxi	0.03	0.25	35	1	4	2	0.5	1	2	0.06	0.25	Not found
153-008	Faropen	1	1	50
	A-C	0.5	0.5	50
	Cefur	0.5	1	50
	Cefdin	0.25	0.25	50
	Azith	1	>64	34	1	0.5	0.5	0.25	32	8	0.016	0.03	Not found
	Teli	1	2	50
	Levo	0.03	0.06	50
	Moxi	0.03	0.06	50
110-061	Faropen	2	2	50
	A-C	2	2	50
	Cefur	2	4	50
	Cefdin	0.5	0.5	50
	Azith	1	16	22	2	1	4	1	16	32	0.03	0.06	Not found
	Teli	2	4	50
	Levo	0.03	0.06	50
	Moxi	0.03	0.25	34	2	1	2	0.5	1	2	0.125	0.5	GyrA(D88N)
HI-42	Faropen	1	2	50
	A-C	1	1	50
	Cefur	2	1	50
	Cefdin	0.5	0.5	50
	Azith	0.5	4	29	1	1	1	0.25	4	4	8	4	Not found
	Teli	2	4	50
	Levo	8	8	50
	Moxi	4	8	50
153-026	Faropen	0.5	1	50
	A-C	0.5	0.5	50
	Cefur	1	2	50
	Cefdin	0.25	0.25	50
	Azith	2	32	24	0.5	0.5	1	0.25	64	>64	0.016	0.03	Not found
	Teli	1	8	49	0.5	0.5	0.5	0.25	2	4	0.016	0.016	Not found
	Levo	0.03	0.06	50
	Moxi	0.03	0.06	50

^aCross-reactivity is denoted in bold. Abbreviations: Faropen, faropenem; A-C, amoxicillin-clavulanate; Cefur, cefuroxime; Cefdin, cefdinir; Azith, azithromycin; Teli, telithromycin; Levo, levofloxacin; Moxi, moxifloxacin.

For single-step mutation studies, approximately 1 × 10¹⁰ CFU/ml in 100-μl aliquots of each strain were spread on sheep blood-Mueller-Hinton agar (pneumococci) or Haemophilus test medium (H. influenzae) plates containing 2×, 4×, or 8× agar dilution MICs of each compound (1, 4, 12, 14). Viability counts and mutation frequencies were calculated as before (14). As described above, the identities of all resistant clones and parents were confirmed by pulsed-field gel electrophoresis.

The MICs (μg/ml) of the parent S. pneumoniae strains (Table 1) were as follows: faropenem, 0.03 to 1; amoxicillin- clavulanate, 0.016 to 2; cefuroxime, 0.03 to 8; cefdinir, 0.03 to 8; azithromycin, 0.016 to 8 (four strains with MICs of >64 μg/ml were not tested); telithromycin, 0.004 to 0.5; levofloxacin, 0.5 to 16; and moxifloxacin, 0.125 to 4. Faropenem, amoxicillin-clavulanate, cefuroxime, and cefdinir did not yield mutants with raised MICs even after 50 days (Table 1). By contrast, azithromycin yielded mutants of three of the five strains tested after 20 to 33 days, with MICs of 0.016 to 0.03 μg/ml (parents) and 0.25 to >64 μg/ml (mutants), and telithromycin gave mutants of seven of nine strains after 14 to 30 days, with parent MICs of 0.004 to 0.5 μg/ml and mutant MICs of 0.016 to >64 μg/ml. Levofloxacin had one strain with a raised MIC, from 1 μg/ml to 16 μg/ml, after 18 days, and moxifloxacin gave mutants of six of nine strains after 19 to 50 days, with parent MICs from 0.125 to 1 μg/ml and mutant MICs from 1 to 8 μg/ml. Changes in GyrA (S81F/Y, E85Q), GyrB (P413S, S478I), and ParC (S79F/Y, D83N) were found in both moxifloxacin and levofloxacin mutants. Alterations in azithromycin and telithromycin mutants were detected in 23S rRNA and L22 protein.

The MICs of parent H. influenzae strains (μg/ml) (Table 2) were as follows: faropenem, 0.5 to 2; amoxicillin-clavulanate, 0.5 to 2; cefuroxime, 0.5 to 2; cefdinir, 0.25 to 1; azithromycin, 0.25 to 2; telithromycin, 0.5 to 2; levofloxacin, 0.03 to 8; and moxifloxacin, 0.03 to 4. Faropenem, amoxicillin-clavulanate, cefuroxime, and cefdinir did not yield mutants with raised MICs even after 50 subcultures. By comparison, azithromycin yielded resistant mutants of all six strains after 15 to 34 days, with 0.25 to 2 μg/ml (parents) and 4 to >64 μg/ml (mutants). Telithromycin gave resistant mutants of three of six strains after 21 to 49 days, with parents at 0.5 to 1 μg/ml and mutants at 4 to 8 μg/ml. Of the two quinolones tested, levofloxacin had mutants of two of six strains after 28 days, with parents at 0.03 μg/ml and mutants at 0.25 to 0.5 μg/ml. Moxifloxacin yielded mutants of three of six strains after 34 to 39 days, with parents at 0.03 to 0.06 μg/ml and mutants at 0.25 to 0.5 μg/ml. For both quinolones, mutants still had MICs in the susceptible range. Changes in GyrA (S84L and D88N) and GyrB (E469D) or ParE (S458L) were found in two of three moxifloxacin and all levofloxacin mutants (Table 2). Alterations in L4 or L22 protein were detected in all telithromycin-selected mutants. In contrast, all azithromycin mutants had no changes in L4, L22, or 23S rRNA, and the mechanisms of resistance have not been defined.

Results for single-step experiments with S. pneumoniae are shown electronically in Table S3 in the supplemental material. The mutation frequencies for faropenem ranged between <2.5 × 10⁻¹⁰ and 1.8 × 10⁻⁹ (2× MIC) and <2.5 × 10⁻¹⁰ to <5.0 × 10⁻¹⁰ (8× MIC); the maximal faropenem MIC of recovered clones was 0.03 μg/ml. The amoxicillin-clavulanate mutation frequencies were <5.0 × 10⁻¹¹ to <4.0 × 10⁻¹⁰ at both 2× and 8× MICs; moxifloxacin frequencies were <1.0 × 10⁻¹⁰ to <1.0 × 10⁻⁸ at both 2× and 8× MICs. The cefdinir mutation frequencies ranged between <1.3 × 10⁻¹⁰ to 6.7 × 10⁻⁷ (2× MIC) and <1.3 × 10⁻¹⁰ to <6.7 × 10⁻¹⁰ (8× MIC). The cefuroxime mutation frequencies were between <2.5 × 10⁻¹⁰ to 1.5 × 10⁻⁸ (2× MIC) and <2.5 × 10⁻¹⁰ to <1.0 × 10⁻⁹ (8× MIC). The telithromycin mutation frequencies were between 1.1 × 10⁻⁹ to 1.3 × 10⁻⁴ (2× MIC) and <1.5 × 10⁻¹⁰ to 4.8 × 10⁻⁶ (8× MIC). For the three azithromycin mutants tested, the mutation frequencies at 2× and 8× MICs were 5.0 × 10⁻⁹ to 7.2 × 10⁻⁹ and <1.9 × 10⁻¹⁰ to <1.3 × 10⁻⁹, respectively. For the seven levofloxacin mutants tested, the mutation frequencies at 2× and 8× MICs were <1.0 × 10⁻¹⁰ to 7.9 × 10⁻⁸ and <1.0 × 10⁻¹⁰ to <2.6 × 10⁻¹⁰, respectively.

Results for single-step experiments with H. influenzae are shown electronically in Table S4 in the supplemental material. The mutation frequencies for faropenem were <7.1 × 10⁻¹¹ to <5.0 × 10⁻¹⁰ for both 2× and 8× MICs. No mutants were selected by the single-step method for faropenem, cefuroxime, or cefdinir. Cefuroxime ranged from <7.1 × 10⁻¹¹ to <5.0 × 10⁻¹⁰ (2× and 8× MICs) and cefdinir from <7.1 × 10⁻¹¹ to <3.3 × 10⁻¹⁰ (2× and 8× MICs). Amoxicillin-clavulanate ranged between <1.7 × 10⁻¹⁰ to 2.1 × 10⁻⁷ (2× MIC) and <1.7 × 10⁻¹⁰ to <5.0 × 10⁻¹⁰ (8× MIC). The azithromycin frequencies were between <3.3 × 10⁻¹⁰ to 2.3 × 10⁻⁸ (2× MIC) and <2.5 × 10⁻¹⁰ to <5.0 × 10⁻¹⁰ (8× MIC). For telithromycin, the mutation frequencies at 2× and 8× MICs were <3.8 × 10⁻¹⁰ to 1.6 × 10⁻⁷ and <2.5 × 10⁻¹⁰ to <3.8 × 10⁻¹⁰, respectively. For moxifloxacin, the mutation frequencies at 2× and 8× MICs were >2.6 × 10⁻¹⁰ to 3.0 × 10⁻⁸ and <1.7 × 10⁻¹⁰ to <5.0 × 10⁻¹⁰, respectively.

In our previous resistance selection studies with S. pneumoniae and H. influenzae (1-4, 7, 12-14), all potent β-lactams with low MICs failed to select for resistance in multistep studies and yielded low frequencies of spontaneous mutations in single-step experiments. Examples are amoxicillin-clavulanate and ceftriaxone against both species and cefixime and cefpodoxime against H. influenzae. In two previous recent studies from our group, imipenem did not yield resistant mutants of pneumococci after 50 subcultures (4, 13), with very low frequencies of spontaneous mutations according to the single-step method (4): these results are in line with the faropenem results in the current study. A preliminary single-step study has shown that imipenem and meropenem both show very low rates of spontaneous mutation, similar to those found for faropenem in our study, against H. influenzae (K. Nagai, unpublished data). We could not find references in the literature to multistep resistance selection of imipenem and meropenem against H. influenzae, and we have not performed such work ourselves. Koga and coworkers (11) found that imipenem did not select for resistant H. influenzae strains in an in vitro pharmacodynamic model. Thus, our current findings with faropenem are in line with those of other potent β-lactam agents against both of these two species. These findings contrast with those of macrolides and especially quinolones, which readily yield resistant mutants with either method (1-4, 7, 12-14) and are potentially important clinically, as mechanisms for resistance in the three drug classes are different.

Faropenem not only gave low MICs against all pneumococcal and H. influenzae resistotypes tested but, similar to other potent β-lactams, failed to select for resistant clones after 50 daily subcultures (unlike macrolides and quinolones) and also yielded very low resistance frequencies in single-step studies. These results indicate a low propensity in faropenem for selection for resistant pneumococci and H. influenzae when introduced clinically.