|Home | About | Journals | Submit | Contact Us | Français|
MICs of linezolid in broth microdilutions were tested against 341 slowly growing nontuberculous mycobacteria (NTM) belonging to 15 species. The proposed linezolid susceptibility MICs for all Mycobacterium marinum, Mycobacterium szulgai, Mycobacterium kansasii, Mycobacterium malmoense, and Mycobacterium xenopi isolates and for 90% of Mycobacterium gordonae and Mycobacterium triplex isolates were ≤8 μg/ml. Linezolid has excellent therapeutic potential against most species of NTM.
Treatment of infections due to slowly growing nontuberculous mycobacteria (NTM) remains difficult for many species. In some cases only a few drugs are available for therapy, and in most situations combination therapy is necessary. Previous in vitro studies with linezolid have shown it to be active against most species of rapidly growing mycobacteria (RGM) (11) and Nocardia (4) and recently against Mycobacterium tuberculosis (1) at readily achievable levels in serum. Thus, we undertook a study of the in vitro activity of linezolid against species of slowly growing NTM.
(This work was presented in part at the 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, Calif., 27 to 30 September 2002.)
We tested linezolid against 341 isolates of slowly growing NTM belonging to 15 species. This number included 335 clinical strains submitted for susceptibility testing and/or identification and six reference isolates kindly provided by the American Type Culture Collection (ATTC). The isolates included those of the Mycobacterium avium complex (MAC) (189 isolates), Mycobacterium marinum (47 isolates), Mycobacterium szulgai (10 isolates), Mycobacterium gordonae (21 isolates), Mycobacterium kansasii (19 isolates), Mycobacterium simiae complex (15 isolates), Mycobacterium terrae complex (11 isolates), Mycobacterium triplex (10 isolates), Mycobacterium lentiflavum (5 isolates), Mycobacterium xenopi (5 isolates), Mycobacterium malmoense (5 isolates), and one isolate each of Mycobacterium interjectum, Mycobacterium asiaticum, Mycobacterium scrofulaceum, and Mycobacterium branderi. Susceptibility was determined once for each isolate except for some reference strains.
The isolates were identified by standard methods, including a combination of traditional biochemicals and high-performance liquid chromatography (2, 6; S. H. Chiu, K. C. Jost, Jr., D. F. Dunbar, and L. B. Elliott, Abstr. 98th Gen. Meet. Am. Soc. Microbiol. 1998, abstr. U-76, p. 508, 1998), nucleic acid probes, (5) or PCR-restriction fragment length polymorphism analysis of the 439-bp Telenti fragment of the 65-kDa hsp gene (9, 10). Most isolates of the MAC were identified by high-performance liquid chromatography and/or nucleic acid probes. All species other than those of the MAC, M. kansasii, M. marinum, and M. gordonae were confirmed by PCR-restriction fragment length polymorphism analysis.
MICs were determined by using NCCLS-recommended serial twofold broth microdilutions in cation-adjusted Mueller-Hinton broth (12). MICs were tested with multiple lots of plates with linezolid concentrations ranging from ≤0.5 to 128 μg/ml. The MIC breakpoints were those proposed by the NCCLS for testing linezolid against RGM (12). Results were determined after 7 days of incubation at 35°C. The end point was complete (100%) inhibition of visible growth.
Quality control assays were performed with Staphylococcus aureus ATCC 29213; the linezolid MIC range for this strain is 1 to 4 μg/ml (after 18 to 24 h of incubation) (8). Additional quality control assays were performed with M. marinum ATCC 927T, M. avium ATCC 700898, M. avium ATCC 35712, M. avium ATCC 35718, M. triplex ATCC 700071T, and M. malmoense ATCC 29571T.
Generally, linezolid showed excellent activity, with most MICs in the proposed ranges for susceptible (MIC, ≤8 μg/ml) or intermediate (MIC, ≤16 μg/ml) organisms (Tables (Tables11 and and2).2). The MICs for five isolates of M. lentiflavum were 8 to 16 μg/ml, and the MICs for five isolates of M. xenopi were 4 to 8 μg/ml (data not shown). The MICs for the single isolates of M. interjectum and M. asiaticum were 16 μg/ml (data not shown).
In contrast, linezolid was less active in vitro against isolates of the MAC, the M. terrae complex, and the M. simiae complex. The MICs for the single isolates of M. scrofulaceum and Mycobacterium branderi were >32 and 32 μg/ml, respectively (data not shown).
The MICs for the S. aureus quality control strain were within the expected range. M. marinum ATCC 927T was tested 58 times; the MIC for this strain was ≤2 μg/ml (58 values). M. avium ATCC 700898, which is the NCCLS-recommended strain for susceptibility testing of clarithromycin with the MAC, was also tested 23 times, with a resulting modal MIC of 32 μg/ml (range, 8 to 32 μg/ml). M. triplex ATCC 700071 was tested twice, with a resulting MIC of 16 μg/ml both times. Other quality control organisms were tested only once. The MICs for M. avium ATCC 35712, M. avium ATCC 35718, M. triplex ATCC 700071T, and M. malmoense ATCC 29571T were 64, 16, 16, and ≤2 μg/ml, respectively.
Previous studies have shown that 90% of the isolates of the RGM species Mycobacterium fortuitum and Mycobacterium chelonae were inhibited by ≤16 μg of linezolid per ml, with modal MICs of ≤8 μg/ml (11). Two immunosuppressed patients with macrolide-resistant disseminated M. chelonae infections as well as small numbers of other, nonimmunosuppressed patients have been successfully treated with linezolid (3; B. A. Brown-Elliott, R. J. Wallace, Jr., D. E. Griffith, D. Lakey, E. Moylett, M. Gareca, T. R. Perry, R. Blinkhorn, and D. Hopper, Abstr. 40th Annu. Meet. Infect. Dis. Soc. Am. 2002, abstr. 609, p. 151, 2002; M. G. Gareca, B. A. Brown-Elliott, and R. J. Wallace, Jr., Abstr. 40th Annu. Meet. Infect. Dis. Soc. Am. 2002, abstr. 265, p. 91, 2002). In the present study, linezolid demonstrated similar activities against most commonly encountered, slowly growing NTM, with only isolates of the MAC, the M. terrae complex, and M. simiae requiring ≥32 μg/ml to inhibit 90% of the isolates tested. M. marinum, M. kansasii, M. gordonae, M. malmoense, and M. szulgai were the most susceptible species, with 100% of the strains inhibited by ≤8 μg of linezolid per ml.
The peak levels of linezolid in serum after oral doses of 600 mg twice daily are 21.2 ± 5.8 μg/ml with a half-life of 5.4 h (package insert for Zyvox, Pharmacia and The Upjohn Co., Kalamazoo, Mich.). This suggests that 16 μg/ml may be a reasonable intermediate value for other organisms such as the NTM. Pending further clinical experience, we propose the following MIC breakpoints for the slowly growing NTM species: for susceptible isolates, ≤8 μg/ml; for intermediate isolates, 16 μg/ml; and for resistant isolates, ≥32 μg/ml. These same breakpoints have also been proposed by the NCCLS for antimycobacterial susceptibility testing (11, 12) for the RGM. For quality control in susceptibility testing, we recommend the use of M. avium ATCC 700898, with an acceptable linezolid MIC range of 8 to 32 μg/ml. We also recommend linezolid concentrations of 2 to 32 μg/ml for testing isolates of slowly growing NTM.
The use of linezolid for treatment of NTM infections has been very limited. This in part reflects the lack of long-term safety data, concern over limiting adverse hematologic events, and the cost of the drug (Brown-Elliott et al., Abstr. Annu. Meet. Infect. Dis. Soc. Am., 2002). One study suggested that once-daily adult dosing (600 mg) rather than the standard 600-mg twice-daily dosing that is used for bacterial species may be adequate for the treatment of mycobacterial infections and may also help to limit bone marrow suppression (Brown-Elliott et al., Abstr. Annu. Meet. Infect. Dis. Soc. Am., 2002). Linezolid has been used successfully to treat small numbers of patients with Nocardia (7) and RGM (3; Brown-Elliott et al., Abstr. Annu. Meet. Infect. Dis. Soc. Am., 2002; Gareca et al., Abstr. 40th Annu. Meet. Infect. Dis. Soc. Am., 2002), infections, which suggests that the drug may work for infections with slowly growing NTM.
We thank Pharmacia Inc. for support of this research.
We also thank Joanne Woodring for preparation of the manuscript and the mycobacteriology section of the Texas Department of Health for assistance with the identification of some of the isolates.