No abstract

Mycobacterium smegmatis is a rapidly growing environmental species not considered a human pathogen. We identified 22 human isolates of M. smegmatis from Australia and the southern United States: 19 were from skin or soft-tissue infections, and none were from urine or the male genital tract. These isolates closely resembled Mycobacterium fortuitum, except for a negative three-day arylsulfatase test; growth at 43-45 C; a low semiquantitative catalase test; and, in 50% of isolates, a late-developing, yellow-to-orange pigment. The isolates were biochemically identical to four reference strains and the type strain of M. smegmatis. Isolates were resistant to isoniazid and rifampin but susceptible to ethambutol, doxycycline, sulfamethoxazole, ciprofloxacin, imipenem, and amikacin. Eleven patients treated on the basis of in vitro susceptibility tests responded well to therapy. The similarity of M. smegmatis to M. fortuitum and the failure to recognize that the former is an environmental species may have contributed to previous failures to recognize it as a human pathogen.

One hundred twenty-three patients with nonpulmonary infections due to Mycobacterium fortuitum or Mycobacterium chelonei were treated by wound debridement and with chemotherapy on the basis of in vitro susceptibilities of the organism. Of 76 patients with infections caused by M. fortuitum, 13 required no therapy or were adequately treated with surgery alone. Patients with active localized disease received single drug therapy (usually with a sulfonamide) for a mean period of 10.6 weeks for cellulitis and seven months for osteomyelitis. Patients with extensive disease received amikacin or amikacin plus cefoxitin (mean, four weeks) followed by a sulfonamide (mean, six months). The 47 patients with infections caused by M. chelonei received no therapy or were treated with surgery alone (6); with amikacin (10), erythromycin (6), doxycycline (3), or cefoxitin (1); or with amikacin plus cefoxitin followed by cefoxitin alone for a total of 10-12 weeks (20); or other multiple-drug regimens (1). Surgery was performed on 74 (60%) patients. Schlichter tests or serum drug levels were determined for 81 (66%) patients. Response to therapy was excellent; 68 (90%) infections with M. fortuitum and 34 (72%) with M. chelonei were successfully treated. Cultures became negative within six weeks of chemotherapy, except for sternal osteomyelitis, for which cultures were not negative until up to 14 weeks. Follow-up for a mean period of 12 months following therapy was possible in 80% of cases. Relapses were rare except in patients with disseminated disease, and drug resistance developed in only one patient. These studies demonstrate the value of routine susceptibility testing of these mycobacterial species and the benefit of chemotherapy on the basis of in vitro susceptibilities.

Previous studies of Mycobacterium fortuitum identified isolates that did not fit its two recognized biovariants. Eighty-five clinical isolates of this group, the "third biovariant complex", were evaluated. They represented 16% of 410 isolates of M. fortuitum submitted to a Texas laboratory and 22% of 45 isolates in Queensland, Australia. Most infections (76%) involved skin, soft tissue, or bone and occurred after metal puncture wounds or open fractures. Isolates differed from biovar fortuitum in resistance to pipemidic acid and use of mannitol and inositol as carbon sources. Two subgroups were present, and examples were deposited in the American Type Culture Collection. Isolates were resistant to doxycycline and one-third were resistant to cefoxitin. All were susceptible to amikacin, ciprofloxacin, sulfamethoxazole, and imipenem. Surgical debridement combined with drug therapy based on in vitro susceptibilities resulted in cures of cutaneous disease or osteomyelitis. DNA homology studies are needed to determine the taxonomic status of these organisms.

A total of 18 strains of Mycobacteriumfortuitum, 15 strains of M. chelonei, and 31 strains of M. chelonei-like organisms were tested by both broth microdilution and agar dilution to determine their susceptibility to 34 antimicrobial agents. All strains grew well enough in cation-supplemented Mueller-Hinton broth for endpoints to be read after 72 h of incubation. Some strains of M. chelonei did not grow on Mueller-Hinton agar. A few discrepancies were noted between the broth and agar procedures. For M. fortuitum, doxycycline, minocycline, amikacin, sulfamethoxazole, and sulfamethoxazole-trimethoprim were the most active agents. For M. chelonei, amikacin, sisomicin, tobramycin, and erythromycin were the most active agents. The M. chelonei-like organisms were most susceptible to ampicillin, doxycycline, minocycline, amikacin, erythromycin, sulfamethoxazole, and sulfamethoxazole-trimethoprim. Broth microdilution appears to be a reliable method for susceptibility testing of rapidly growing mycobacteria, although clinical studies are needed to determine how well in vitro results correlate with therapeutic in vivo outcome.

Broth microdilution MICs were determined for 258 clinical isolates of Mycobacterium fortuitum (3 biovariants) and M. chelonae (2 subspecies) with amikacin, tobramycin, cefoxitin, doxycycline, erythromycin, and sulfamethoxazole-trimethoprim and with several new jI-lactams and aminoglycosides and ciprofloxacin. Variations in susceptibility by and within species subgroups confirm the need for susceptibility testing against clinically important strains.The role of the rapidly growing mycobacteria Mycobacterium fortuitum and M. chelonae (formerly M. chelonei) as agents of human disease has been clearly established (22). Except for the aminoglycosides, these organisms are resistant to the antituberculosis agents, and testing against these drugs is no longer indicated (19). However, they are susceptible to some drugs not routinely tested in laboratories specializing in mycobacteriology; these agents have been shown to be effective in the treatment of human disease (8, 9, 19, 21a). Because the susceptibilities of M. fortuitum and M. chelonae to these agents varies, testing clinically important strains as an aid in determining a therapeutic regimen is important. Although the need to identify organisms of the M. fortuitum complex to subspecies and even species level remains controversial (11,17), previous studies have suggested that patterns of susceptibility may relate in part to differences between species, subspecies, and biovariants (17,19,20,21,23,24).Between 1978 and 1982, more than 250 strains of rapidly growing mycobacteria were received in our laboratories for identification and susceptibility testing. In this report, we present the antimicrobial susceptibilities of organisms that were identified to the subspecies or subgroup level as determined by a broth microdilution method. Our results provide susceptibility data on a large number of organisms and confirm the need to determine the susceptibilities of M. fortuitum and M. chelonae. In addition, the results show that each species subgroup has a fairly typical pattern of susceptibility.MATERIALS AND METHODS Ot-ganisms. Isolates were submitted to our laboratories for either identification or antimicrobial susceptibility determination, or both. Organisms were identified by standard methods (17), except that the unnamed third biovariant of M. fortuitum was expanded to a complex which includes isolates positive for two or more carbohydrates (mannitol, inositol, or citrate). The strains and number included were M. fortuitum biovariant fortuitum, 98; M. fortuitum biovariant peregrinum, 8; M. fortuitum third biovariant complex, 19; M. chelonae subsp. abscessus, 99; and M. chelonae subsp. chelonae, 34. A small number of these strains were included in our previous study (19). Isolates of the M. chelonae-like * Corresponding author. group were not included in this report because results with most of these strains were reported earlier (9). Aspects of clinical disease and results of therapy for many of these isolates have been reported (8, 9, 21a, 22).The organisms were st...

Eighty-nine isolates of rapidly growing mycobacteria associated with cardiac bypass-related infections were characterized. Isolates from sporadic infections belonged to eight taxonomic groups and displayed numerous multilocus enzyme genotypes, plasmid profiles, and heavy metal and antibiotic resistance patterns. Compared with 449 noncardiac wound isolates, 45 sporadic cardiac isolates were more likely to be Mycobacterium fortuitum and M. smegmatis and less likely to be M. chelonae. About 80% of cardiac and noncardiac isolates were from southern coastal states. Eight outbreaks of cardiac bypass-related infections were identified. Strains from each outbreak were genotypically distinctive, and five outbreaks involved more than one strain. In two outbreaks, isolates from environmental sources and noncardiac infections were similar or identical to isolates from sternal wound infections. The heterogeneity of these isolates suggests that most isolates are unrelated and are derived from local environmental sources rather than from contaminated commercial surgical materials or devices.

Between April and November 1982, 27 of 140 patients in a hemodialysis center in Louisiana were infected with rapidly growing mycobacteria; 14 had bacteremia alone, 3 had soft-tissue infections, 1 had an access-graft infection, and 9 had widely disseminated disease. Of 26 identified isolates, 25 were Mycobacterium chelonei ssp. abscessus, and one was an M. chelonei-like organism. One factor common to all patients was exposure to processed hemodialyzers (artificial kidneys). Environmental sampling of the water-treatment system showed widespread contamination with nontuberculous mycobacteria, which were also recovered from the patient's side (blood compartment) of five of 31 hemodialyzers that had been processed and were ready for use. The formaldehyde concentration was less than 2% in two of three such contaminated dialyzers tested. We hypothesize that patients became infected when their blood circulated through processed dialyzers that contained viable rapidly growing mycobacteria. This outbreak demonstrates that hemodialysis patients may be at risk for developing infections with rapidly growing mycobacteria and that such infections may go unrecognized when routine culture methods are used. It also emphasizes the importance of using effective procedures to disinfect dialyzers in hemodialysis centers.