Determination of epidemic clonality among multidrug-resistant strains of Acinetobacter spp. and Pseudomonas aeruginosa in the MYSTIC Programme (USA, 1999–2003)

Abstract

The Meropenem Yearly Susceptibility Test Information Collection (MYSTIC) Programme was initiated in 1997 (1999 for the United States). This program monitors resistance in participant medical centers where carbapenems are prescribed and drug use data can be obtained. An earlier report found antimicrobial use was not a clear cause of local or aggregate changes in resistance rates. This study addresses the role of dissemination of resistant clones on susceptibility rates for nonfermentors, Acinetobacter spp. (ACB) and Pseudomonas aeruginosa (PSA). Carbapenem (CARB)-multidrug-resistant strains (MDR) from among 236 ACB and 1,111 PSA were tested by reference broth microdilution methods, automated ribotyping, and pulsed field gel electrophoresis to determine possible clonal dissemination. Each strain was also tested for metallo-β-lactamases (MβL) (phenotypic and polymerase chain reaction); and then analyzed by CARB-R rate and defined daily dose (DDD)/100 days use groupings (high, moderate, and low). For the aggregate 15 sites in the MYSTIC Programme each year, the CARB-resistant rate decreased over 5 years; but other drug-resistance rates generally escalated. Changes were not related to antimicrobial use calculations. The discovered clonally spread MDR-PSA strains were more frequent in high- (1.8 clones/site) and moderate-resistance (0.6 clones/site) rate centers (21.7% to 29.5% were clonal), compared with unique strains in low-resistance hospitals. ACB clonality was extreme in one geographic area, with dissemination of 5 different clones (931.7/B, C, or D; 1090.2/A; 167.5/A) in 4 centers (02, 04, 06, and 18). Resistance rates in ACB and PSA were clearly related to clonal occurrence and spread, and one MβL (VIM-7) was detected. Decreased CARB resistance rates from 1999 through 2002 were directly attributed to the disappearance of resistance clones in some locations. In conclusion, ACB and PSA CARB and MDR resistance rates in MYSTIC Programme institutions have been greatly influenced by clonal dissemination and less by antimicrobial use patterns. The most serious examples of resistance were the clonality observed among ACB in New York City and the documented endemic nature of VIM-7-producing PSA (0.09% of all PSA isolates). Meropenem remained the most active antimicrobial agent tested in the program, and surveillance networks must implement epidemiologic typing to accurately assess the role of clonal spread on the study results.