In Vitro Comparison of Ertapenem, Meropenem, and Imipenem against Isolates of Rapidly Growing Mycobacteria and Nocardia by Use of Broth Microdilution and Etest

Abstract

We compared the activities of the carbapenems ertapenem, meropenem, and imipenem against 180 isolates of rapidly growing mycobacteria (RGM) and 170 isolates of Nocardia using the Clinical and Laboratory Standards Institute (CLSI) guidelines. A subset of isolates was tested using the Etest. The rate of susceptibility to ertapenem and meropenem was limited and less than that to imipenem for the RGM. Analysis of major and minor discrepancies revealed that >90% of the isolates of Nocardia had higher MICs by the broth microdilution method than by Etest, in contrast to the lower broth microdilution MICs seen for >80% of the RGM. Imipenem remains the most active carbapenem against RGM, including Mycobacterium abscessus subsp. abscessus. For Nocardia, imipenem was significantly more active only against Nocardia farcinica. Although there may be utility in testing the activities of the newer carbapenems against Nocardia, their activities against the RGM should not be routinely tested. Testing by Etest is not recommended by the CLSI.

INTRODUCTION

Treatment of infections due to rapidly growing mycobacteria (RGM) and Nocardia remains difficult in part because of resistance to first-line antituberculous agents (for RGM) and other antimicrobial agents (for both RGM and Nocardia) (1, 2). Previous studies with the carbapenems have shown that these agents have limited activity against most pathogenic RGM, but few data on their activity against Nocardia exist (3, 4). Although ertapenem and meropenem have been in use against clinically significant bacteria for several years, there has been a paucity of data on the activities of these agents against RGM and Nocardia. Thus, we undertook a comparative study of the in vitro susceptibilities to imipenem, meropenem, and ertapenem of the most commonly encountered species of RGM and Nocardia, including the Mycobacterium fortuitum group (M. fortuitum, M. senegalense, M. porcinum); M. chelonae; the M. abscessus complex, including M. abscessus subsp. abscessus (formerly M. abscessus and here referred to as M. abscessus), M. abscessus subsp. bolletii (here referred to as M. bolletii), and M. abscessus subsp. massiliense (here referred to as M. massiliense); the M. mucogenicum/M. phocaicum group; M. neoaurum; M. goodii; M. immunogenum; Nocardia cyriacigeorgica; members of the N. nova complex; N. farcinica; N. brasiliensis; N. abscessus; N. otitidiscaviarum; members of the N. transvalensis complex; and N. pseudobrasiliensis. The taxonomy of the M. abscessus complex is currently controversial (5). However, for clarity, we have chosen to use the taxonomy proposal made prior to 2011 to combine the species M. massiliense and M. bolletii into one subspecies (i.e., M. abscessus subsp. bolletii). We also compared MICs for selected isolates from several of the clinically significant species using the Clinical Laboratory and Standards Institute (CLSI)-recommended broth microdilution method and the Etest.

MATERIALS AND METHODS

Organisms.

Clinical isolates of RGM and Nocardia submitted to the University of Texas Health Science Center at Tyler, TX (UTHSCT), for susceptibility testing from 2006 to 2008 were selected for testing. This set included 180 isolates of RGM (67 M. abscessus isolates, 11 M. massiliense isolates, 3 M. bolletii isolates, 38 M. fortuitum isolates, 10 M. porcinum isolates, 7 M. senegalense isolates, 21 M. chelonae isolates, 16 M. mucogenicum/M. phocaicum group isolates, and 7 other isolates of RGM, including 2 isolates of the M. neoaurum-M. lacticola group, 3 M. goodii isolates, and 2 M. immunogenum isolates). The 170 isolates of Nocardia tested included 26 N. cyriacigeorgica isolates, 57 N. nova complex isolates, 13 N. abscessus isolates, 23 N. brasiliensis isolates, 19 N. farcinica isolates, 18 N. transvalensis complex isolates, 8 N. otitidiscaviarum isolates, 1 N. pseudobrasiliensis isolate, and 5 Nocardia sp. isolates.

Isolates of RGM and Nocardia were identified to the species level by molecular methods, including PCR restriction enzyme analysis (PRA) of a 441-bp sequence of the 65-kDa hsp gene (6, 7), and their antimicrobial susceptibility patterns (2, 8–15). Isolates that were not identifiable by PRA were subjected to 16S rRNA (16) and/or multigene target (hsp65, secA1, rpoB, etc.) sequence analysis (3, 17–20).

Susceptibility testing.

The MICs of imipenem, meropenem, and ertapenem were determined by broth microdilution using the CLSI-recommended procedure and interpretive criteria for RGM and Nocardia (21) with imipenem. MIC panels were custom manufactured to include meropenem and ertapenem by Thermo Fisher (formerly Trek Diagnostics, Inc.). Since no interpretive criteria for mycobacteria and Nocardia with ertapenem are currently available, the CLSI-recommended intermediate (I) breakpoint (4 μg/ml) for testing of bacteria was employed (8). These bacterial breakpoints were also applied to testing of the Nocardia with meropenem and ertapenem, which has not been addressed by the CLSI. Additional testing by Etest (AB Biodisk, Uppsala, Sweden) was performed as previously described (22, 23), and the Etest results were compared to the broth microdilution results for 102 selected isolates of RGM and 87 isolates of Nocardia spp. For Etest MICs that fell between doubling dilutions, the results were rounded up to the next 2-fold value, as recommended by the manufacturer. For both methods, results were read at 100% inhibition.

Quality control.

Quality control for the carbapenems was performed by using M. peregrinum ATCC 700686, Enterococcus faecalis ATCC 29212, Staphylococcus aureus ATCC 29213, and Pseudomonas aeruginosa ATCC 27853. CLSI-recommended quality control ranges were those shown in Table 1. Quality control of the broth microdilution assays and Etests was performed at the time of performance of each test.

TABLE 1.

Acceptable MIC ranges^a and numbers of quality control tests performed within those ranges

Quality control strain	Imipenem		Meropenem		Ertapenem
	Acceptable MIC range (μg/ml)	No. of tests	Acceptable MIC range (μg/ml)	No. of tests	Acceptable MIC range (μg/ml)	No. of tests
Enterococcus faecalis ATCC 29212	0.5–2	225/225	2–8	225/225	4–16	225/225
Mycobacterium peregrinum ATCC 700686	2–16	250/250	2–16	250/250	NAb	NAb
Staphylococcus aureus ATCC 29213	0.015–0.06	34/34	0.03–0.12	34/34	0.06–0.25	31/31
Pseudomonas aeruginosa ATCC 27853	1–4	84/84	0.25–1	32/32	2–8	84/84

^aAcceptable MIC ranges have been described previously (8, 21).

^bNA, not available.

RESULTS

Quality control.

All quality control values for all three carbapenems were within acceptable limits with all reference isolates tested (Table 1).

Broth microdilution MICs.

The susceptibilities of the most commonly encountered pathogenic species of RGM and Nocardia to the three carbapenems are shown in Tables 2 and 3, respectively. For the most commonly encountered RGM, the best activity of the carbapenems was noted against isolates of the M. fortuitum group and the M. mucogenicum/M. phocaicum group. Imipenem, meropenem, and ertapenem were active against 100% of 16 isolates of the M. mucogenicum/M. phocaicum group. The MICs of imipenem (MIC₅₀ = 4 μg/ml) and meropenem (MIC₅₀ = 8 to 16 μg/ml) were ≤16 μg/ml for 100% of the isolates in the M. fortuitum group (including M. senegalense and M. porcinum) (Table 2). Only 11% (1/9) of the isolates of M. porcinum had MICs indicating that they were susceptible (S)/intermediate (I) to ertapenem, similar to the results for both M. fortuitum and M. senegalense (0% of which were S/I) (data not shown).

TABLE 2.

Comparison of MIC ranges, MIC₅₀s, MIC₉₀s, and percentage of isolates susceptible/intermediate to imipenem, meropenem, and ertapenem for RGM isolates by broth microdilution

Complex or species and drug	Intermediate breakpoint (μg/ml)	No. of isolates tested	MIC (μg/ml)			% susceptible/intermediate
			Range	50%	90%
M. fortuitum
Imipenem	8–16	38	2–8	4	8	100
Meropenem	8–16	38	4–16	8	16	100
Ertapenem	4a	38	8–32	16	>32	0
M. porcinum
Imipenem	8–16	10	2–16	4	8	100
Meropenem	8–16	9	2–16	16	16	100
Ertapenem	4a	9	1–16	8	16	11
M. abscessus subsp. abscessus
Imipenem	8–16	67	4–>16	>16	>16	66
Meropenem	8–16	67	8–>16	>16	>16	12
Ertapenem	4a	67	8–>32	>32	>32	0
M. massiliense
Imipenem	8–16	11	8–>16	>16	>16	73
Meropenem	8–16	11	≥16	>16	>16	9
Ertapenem	4a	11	≥32	>32	>32	0
M. chelonae
Imipenem	8–16	21	8–16	16	≥16	52
Meropenem	8–16	21	>16–>32	>16	>16	0
Ertapenem	4a	21	>16–>32	>32	>32	0
M. mucogenicum/M. phocaicum group
Imipenem	8–16	16	≤0.5–4	2	4	100
Meropenem	8–16	16	≤0.5–8	4	8	100
Ertapenem	4a	16	2–4	2	4	100

^aBased on the CLSI breakpoint for bacteria.

TABLE 3.

Comparison of ranges, MIC₅₀s, MIC₉₀s, and percentage of isolates susceptible/intermediate to imipenem, meropenem, and ertapenem for Nocardia isolates by broth microdilution

Complex or species and drug	Intermediate breakpoint (μg/ml)	No. of isolates tested	MIC (μg/ml)			% susceptible/intermediate
			Range	50%	90%
N. cyriacigeorgica
Imipenem	8	25	≤1–32	8	>16	60
Meropenem	8	25	4–>16	8	>16	68
Ertapenem	4a	26	2–>16	8	>16	15
N. nova complex
Imipenem	8	57	≤0.5–8	≤1	2	100
Meropenem	8	54	≤0.5–16	≤1	4	94
Ertapenem	4a	57	0.5–16	2	4	96
N. abscessus
Imipenem	8	13	2–32	>16	32	23
Meropenem	8	11	1–8	2	4	100
Ertapenem	4a	13	0.5–4	2	4	100
N. brasiliensis
Imipenem	8	23	16–>32	>16	>32	0
Meropenem	8	23	4–>16	>16	>16	48
Ertapenem	4a	23	4–>16	>16	>16	26
N. farcinica
Imipenem	8	19	≤1–>16	8	>16	63
Meropenem	8	18	4–>16	8	>16	33
Ertapenem	4a	19	4–>16	8	16	21
N. transvalensis complex
Imipenem	8	18	4–>32	16	>32	22
Meropenem	8	18	2–16	8	16	83
Ertapenem	4a	18	2–>16	>16	>16	22

^aBased on the CLSI breakpoint for bacteria (8).

Other clinically significant but less commonly encountered species included two isolates of the M. neoaurum-M. lacticola group and three isolates of M. goodii, which had 100% susceptibility to all three carbapenems (data not shown). In contrast, the ertapenem MICs for two isolates of M. immunogenum were >32 μg/ml (data not shown). The activity of meropenem was tested against only one isolate of M. immunogenum (MIC > 16 μg/ml); one isolate of M. immunogenum was resistant (R) (MIC > 32 μg/ml) and one was I (MIC = 16 μg/ml) to imipenem.

Sixty-six percent of 67 isolates of M. abscessus and 73% of 11 isolates of M. massiliense had imipenem MICs indicating that they were S/I, whereas meropenem had activity against ≤12% and ertapenem had activity against 0% of isolates in both groups. Meropenem and ertapenem showed no activity against three isolates of M. bolletii, in contrast to the I imipenem MICs (data not shown). Among 21 isolates of M. chelonae, 52% (11/21) had S/I imipenem MICs, and all isolates were R to meropenem and ertapenem, with MIC₅₀s of >16 μg/ml (Table 2).

For the Nocardia, only the members of the N. nova complex had MIC₉₀s in the S range for all three carbapenems (Table 3). One hundred percent of the isolates of the N. abscessus complex exhibited S/I meropenem MICs (11/11) and S/I ertapenem MICs (13/13), but only 23% were S/I to imipenem. The only other taxon against which any carbapenem had >80% activity was the N. transvalensis complex, with 83% (15/18) having S/I meropenem MICs (MIC₅₀ = 8 μg/ml); in contrast, only 22% (4/18) had S/I imipenem and ertapenem MICs. The MIC₉₀s of all three carbapenems for all other species of Nocardia were ≥16 μg/ml. Only 48% (11/23) and 26% (6/23) of the isolates of N. brasiliensis were S/I to meropenem and ertapenem, respectively. One isolate of N. pseudobrasiliensis, eight isolates of the N. otitidiscaviarum complex, and three isolates unable to be identified to the species level (Nocardia spp.) had MIC₉₀s in the R interpretive category for all three carbapenems (data not shown).

Comparison of broth microdilution and Etest MICs.

The MICs for a total of 197 isolates of RGM (n = 102) and Nocardia (n = 95) obtained by the broth microdilution and Etest methods were compared. Table 4 provides a comparison of the very major, major, and minor errors in each taxon studied with isolate numbers of ≥10. The CLSI defines very major errors (VME) to be an interpretive category change from R by the reference method (i.e., in this case, broth microdilution) to S by the method being evaluated (i.e., Etest). A major error is defined as an interpretive category change from S by broth microdilution to R by Etest. Minor errors are those in which one result is I and the other is S or R. In general, most discrepancies were considered minor (interpretive category change from S to I or vice versa or R to I or vice versa).

TABLE 4.

Comparison of Etest MICs and broth microdilution MICs of imipenem, meropenem, and ertapenem for isolates of RGM and Nocardia

Species or complex and drug	No. of isolates tested	% error
		Very major	Major	Minor
M. fortuitum
Ertapenem	22	5	0	0
Meropenem	22	0	0	86
Imipenem	22	0	5	36
M. abscessus subsp. abscessus
Ertapenem	44	0	0	0
Meropenem	44	0	0	14
Imipenem	44	0	7	48
M. chelonae
Ertapenem	13	0	0	0
Meropenem	12	0	0	0
Imipenem	12	0	0	75
N. cyriacigeorgica
Ertapenem	10	0	0	10
Meropenem	10	0	0	20
Imipenem	11	0	36	18
N. nova complex
Ertapenem	32	0	3	43
Meropenem	31	0	0	9
Imipenem	32	0	0	0
N. brasiliensis
Ertapenem	13	0	0	31
Meropenem	13	0	8	31
Imipenem	13	0	0	8
N. transvalensis complex
Ertapenem	12	33	8	17
Meropenem	12	8	8	42
Imipenem	12	8	0	25

The discrepancies between the broth microdilution and Etest susceptibilities for the RGM showed rare (≤7%) major errors in the three major groups (M. fortuitum, M. abscessus, M. chelonae) (Table 4), except for M. chelonae, 75% of 12 isolates of which tested showed minor errors. Only one of five (20%) isolates of M. porcinum showed major discrepancies with ertapenem (data not shown). In all but one case (M. fortuitum), the broth microdilution MICs indicated resistance whereas the Etest reads indicated higher susceptibility than the broth MICs.

Only 1 of 22 isolates of M. fortuitum had a very major error with ertapenem, with the broth microdilution MIC indicating resistance at 8 μg/ml but the Etest indicating susceptibility at 1 μg/ml. Among six isolates of the M. mucogenicum/M. phocaicum group, one isolate had a major error with meropenem (data not shown).

Minor errors were most commonly seen in M. fortuitum with meropenem and imipenem (86% and 36%, respectively). For 15 of 16 isolates (94%), meropenem broth microdilution MICs were 8 μg/ml (intermediate), but Etest MIC reads were ≥16 μg/ml (resistant). Similarly, 3 of 44 isolates of M. abscessus had major errors with imipenem, in which the broth microdilution MICs were ≤4 μg/ml (S) but the Etest MICs were read to be >32 μg/ml (R). Forty-eight percent and 14% of the M. abscessus isolates had minor errors with imipenem and meropenem, respectively. One hundred percent of four isolates of the related group, the M. massiliense group, had major errors with imipenem, and 25% of the same group had minor errors with meropenem. One of four M. senegalense isolates had a broth microdilution imipenem MIC of 8 μg/ml but a susceptible Etest MIC read of 0.5 μg/ml (data not shown). All of the 13 isolates of M. chelonae were resistant by broth microdilution and Etest, and none of the isolates exhibited any minor or major errors with meropenem and ertapenem, but 75% had minor errors with imipenem. In general, ≥80% of the errors (both major and minor) were due to lower broth microdilution MICs rather than lower Etest MICs.

Of the 95 isolates of Nocardia compared by Etests, the majority of very major and major errors with ertapenem were seen with the N. transvalensis complex (Table 4). Of 12 isolates of the N. transvalensis complex tested, 4 isolates (33%) exhibited very major errors with ertapenem and 8% of the isolates showed very major errors with each of meropenem and imipenem. Eight percent of the 12 isolates showed major errors with each of ertapenem and meropenem, while 42% and 25% had minor errors with meropenem and imipenem, respectively. There were also major errors for 4 of 11 (36%) isolates of N. cyriacigeorgica with imipenem. In all four isolates, the broth microdilution MIC indicated resistance (>16 μg/ml), whereas the Etest MICs were ≤4 μg/ml.

For the N. nova complex, there were only rare major errors (1/32, or 3%) for ertapenem. Again, the broth microdilution MICs were higher than the Etest MIC reads.

For the one major error noted with 13 isolates of N. brasiliensis with meropenem, the broth microdilution MIC was susceptible (4 μg/ml), whereas the Etest MIC reading was resistant (16 μg/ml).

Analysis of both the major and minor errors for the nocardiae revealed that >90% of the isolates of Nocardia had higher MICs by broth microdilution than by the Etest. This finding was in contrast to the lower broth microdilution MICs seen for >80% of the RGM.

A comparison of the discrepant results obtained with taxa with >10 isolates tested by broth microdilution and Etest is seen in Table 4. There were significant discrepant results with several species of RGM and Nocardia, but in all cases, the numbers of tests performed by Etest were less than those performed by broth microdilution. Strikingly, different results by both methods were primarily seen with imipenem and were less commonly seen with meropenem. By broth microdilution, 100% of the isolates of M. fortuitum were S/I to all three carbapenems. However, by Etest only, 77%, 27%, and 5% of the isolates were S/I to imipenem, meropenem, and ertapenem, respectively. Another obvious discrepancy was noted with susceptibility to meropenem among isolates of M. senegalense, 100% of which were S/I by broth microdilution but only 25% were S/I by Etest, although the number of isolates compared was less than 10 (data not shown). Similarly, 100% of the nine isolates of M. porcinum tested were S/I to meropenem by broth microdilution, whereas only 33% were S/I to meropenem by Etest. Only 11% of the isolates of this species were S/I to ertapenem by broth microdilution, whereas 50% were S/I to ertapenem by Etest (data not shown). Of the isolates of M. abscessus and M. massiliense tested, 66% and 73%, respectively, were susceptible to imipenem by broth microdilution, whereas 0% were susceptible by Etests (data not shown for M. massiliense). Additionally, 100% of the isolates of the M. mucogenicum/M. phocaicum group showed susceptibility to meropenem by broth microdilution, whereas only 67% showed susceptibility by the Etest method.

Among the Nocardia, the most striking discrepancies were again mostly with meropenem and imipenem. Fifty-four, 68, and 15% of the N. cyriacigeorgica isolates were S/I to imipenem, meropenem, and ertapenem, respectively, by broth microdilution, whereas 0, 36, and 55% were susceptible to the same agents, respectively, by Etest. Another difference was noted with imipenem and ertapenem and isolates of the N. transvalensis complex. By broth microdilution, 22% of the isolates were S/I to both carbapenems; in contrast, 42% were S/I to both carbapenems by Etests. Although only eight isolates were tested, the MICs of meropenem for isolates of the N. otitidiscaviarum complex showed a wide discrepancy, with 43% of isolates being S/I by Etest but only 25% being S/I when broth microdilution was performed (data not shown). Likewise, 48% of the isolates of N. brasiliensis were S/I to meropenem by microdilution, whereas only 23% were S/I to meropenem by Etest. Isolates of N. farcinica were more susceptible to meropenem and ertapenem (33% and 21%, respectively) when they were tested by the broth microdilution method than when they were tested by Etest (both only 13%), although the MICs of imipenem were equivalent (63% of isolates were S/I to imipenem) by both methods.

DISCUSSION

Treatment of infections due to RGM and Nocardia is often difficult because of the lack of antimicrobials with activity against these species. Additional complications arise due to the need for injectable antibiotics for most serious infections. Imipenem has been useful for the treatment of infections caused by most common pathogenic species of RGM and Nocardia, although some species of Nocardia, including N. abscessus and N. brasiliensis, and some isolates of M. chelonae and the M. abscessus complex are resistant. The necessity for the administration of imipenem two to three times daily creates problems for long-term therapy, which is required for the treatment of infections with RGM and Nocardia.

The results of this study indicate that neither the broth microdilution nor Etest MICs of imipenem are able to consistently predict susceptibility or resistance to meropenem and ertapenem. This fact was illustrated in this study with isolates of M. abscessus, in which imipenem had activity against 66% (44 of 67) of the isolates but neither meropenem nor ertapenem had significant activity by both the broth microdilution and Etest methods. Furthermore, although imipenem and meropenem showed activity against all isolates of the M. fortuitum group by broth microdilution, ertapenem was not active against any of these isolates.

For the Nocardia spp., only the isolates of the N. nova complex were uniformly susceptible or intermediate to all three carbapenems. Although isolates of the N. abscessus group were typically resistant to imipenem (10/13, or 77%), interestingly, 100% were S/I to meropenem and ertapenem. Among the isolates of the N. transvalensis complex, meropenem was the most active carbapenem by broth microdilution (22% [4/18] were S/I to ertapenem and imipenem, whereas 83% [15/18] were susceptible to meropenem).

Previous studies have demonstrated the instability of imipenem and meropenem related to the prolonged incubation (greater than 3 to 4 days) sometimes required by isolates of mycobacteria and Nocardia (24). Meropenem was also previously noted to be more stable, with an approximately 50% reduction in activity of the agent at 24 h, in comparison to an 85% loss of activity of imipenem at 24 h (24). This instability likely contributes to the high in vitro MICs seen in susceptibility testing of the carbapenems with these organisms. However, a practical solution for the testing of these agents in the laboratory has not been developed. No similar studies have been performed to test the stability of ertapenem.

Also intriguing was the fact that 48% (11/23) of the isolates of N. brasiliensis were S/I to meropenem, whereas they were completely resistant to imipenem and only marginally susceptible to ertapenem (6/23, or 26%) (Table 2). These results suggest the possibility of some therapeutic potential for meropenem and/or ertapenem against infections involving some groups of Nocardia (N. cyriacigeorgica, N. nova complex, N. abscessus, N. transvalensis complex, and N. brasiliensis). Although less therapeutic potential for these newer carbapenems against isolates of N. brasiliensis and N. cyriacigeorgica exists, the percentage of isolates S/I to meropenem (11/23 [48%] and 17/25 [68%], respectively) may indicate a possible alternative treatment, especially in serious infections with these groups. For isolates of N. farcinica and the N. otitidiscaviarum complex, the most active carbapenem was imipenem. Less than 6/18, or 35%, of the isolates of N. farcinica were S/I to meropenem and ertapenem and only 2/8 (25%) of the isolates of N. otitidiscaviarum were S/I to meropenem and ertapenem. The results of the current study are in concordance with those of a previous Japanese study of the MICs of imipenem and meropenem for these species (25). One possible explanation, according to Sato et al., for this difference in activity between imipenem and meropenem is the presence of a β-lactamase which inactivates imipenem in both N. brasiliensis and N. otitidiscaviarum (26).

Previous large-scale studies (3) focused on susceptibility testing results obtained by the broth microdilution method with imipenem, meropenem, and ertapenem with RGM but did not analyze the MICs of these agents against the Nocardia or differentiate the newly described species or subspecies of RGM (i.e., M. massiliense or M. bolletii, M. porcinum, and M. senegalense). Importantly, imipenem remains the carbapenem of choice for the treatment of infections due to M. abscessus, M. massiliense, and M. chelonae.

Testing of both the RGM and Nocardia by Etest was problematic, with hazy partial zones of inhibition (heavier marginal growth with lighter growth of inside colonies) that were difficult to interpret being detected. A similar observation has recently been reported by Chihara and colleagues when testing isolates of M. abscessus by Etest (27). In the current study, partial zones of inhibition were most often seen with imipenem with more susceptible isolates, such as the M. mucogenicum/M. phocaicum group, M. fortuitum group, N. nova complex, and N. cyriacigeorgica, although some resistant isolates, such as N. abscessus isolates, also posed difficult interpretations when Etest MICs were compared to broth microdilution MICs. Further investigation of the susceptibility to the carbapenems using Etests appears to be warranted before specific recommendations can be made.

Although imipenem has been the carbapenem most commonly used for the treatment of both mycobacterial and nocardial infections, the option of once daily administration of ertapenem makes this newer carbapenem an attractive alternative (28). However, this study indicates that appropriate MIC testing is necessary to ascertain specific susceptibility before meropenem or ertapenem is administered to ensure that the treatment regimen is effective. These studies also suggest that the usage of meropenem should be limited to the treatment of infections due to the M. fortuitum group and the M. mucogenicum/M. phocaicum group. Previous studies have indicated that meropenem has good penetration in lung, bronchial mucosa, and pleural tissues and may thus be useful in serious infections involving these species (29).

There is a paucity of laboratory and clinical data from studies with ertapenem and meropenem. However, a recent case of M. fortuitum infection in a surgical wound of a patient undergoing tendon repair surgery was successfully treated with a combination regimen of clarithromycin, trimethoprim-sulfamethoxazole, and ertapenem for 6 months. Unfortunately, no details, including the laboratory identification method or susceptibility to the newer carbapenems, were published (30).

Among the most commonly encountered Nocardia spp., meropenem showed greater in vitro activity against the N. nova complex, N. abscessus, and the N. transvalensis complex than other groups of Nocardia. Except for infections involving the N. nova complex and N. abscessus, this study suggests that patients should be treated with meropenem only if in vitro testing shows susceptibility to meropenem. Moreover, on the basis of the findings of this study, the use of ertapenem should be considered only with isolates of the M. mucogenicum/M. phocaicum group, the N. nova complex, and the N. abscessus group unless susceptibility testing shows that ertapenem has in vitro activity against these organisms. Thus, larger studies and comparative clinical data from studies with meropenem and ertapenem are needed to make further treatment recommendations for infections caused by RGM and Nocardia.