Inducible clindamycin resistance and molecular epidemiologic trends of pediatric community-acquired methicillin-resistant Staphylococcus aureus in Dallas, Texas

Abstract

Community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) infection occurs commonly in children. Clindamycin resistance may be inducible or constitutive, and the rates of inducible resistance in CA-MRSA that could produce clindamycin treatment failures vary worldwide. The double-disk test was performed in 197 erythromycin-resistant and clindamycin-susceptible CA-MRSA strains from children in Dallas, Texas, from 1999 to 2002 to determine inducible clindamycin resistance. Resistance mechanisms were studied by PCR; epidemiologic trends were studied by pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST). Inducible resistance was demonstrated in 28 (93%+/-6%) of 30 tested isolates in 1999, 21 (64%, +/-11%) of 33 in 2000, 12 (23%+/-7%) of 52 in 2001, and 6 (7%+/-3%) of 82 in 2002. All noninducible strains had the msr(A) gene. Among inducible resistant strains, 31 had erm(B), 24 had erm(C), and 12 had erm(A) genes. Two distinct pulsed types were the most prevalent; one of them was the most common pulsed type in 1999, whereas in 2002 a different pulsed type was prevalent. MLST analyses determined that ST-8 was the most common type, with 76%+/-5% found in 2002. All but one of these clindamycin-susceptible, erythromycin-resistant ST-8 strains showed no induction of clindamycin resistance. We conclude that, among erythromycin-resistant, clindamycin-susceptible CA-MRSA strains isolated from children in Dallas, inducible methylase resistance became less common from 1999 to 2002 (P<0.001). The phenotype of strains was associated with their sequence type. Our results demonstrate a clonal shift in CA-MRSA in Dallas children from 1999 to 2002.

FIG. 1.

Inducible resistance to clindamycin in pediatric CA-MRSA isolates. The proportion of D-test-positive isolates showed a significant linear downward trend for each consecutive year from 1999 to 2002 (P < 0.001; chi-square test for trend analysis).

FIG. 2.

PFGE patterns of common pediatric CA-MRSA clones. Lane 1, PT-F; lane 2, PT-B; lane 3, PT-A.

FIG. 3.

PFGE-defined common clones an d their MLST sequence types in pediatric CA-MRSA isolates, Children's Medical Center of Dallas, 1999 to 2002. One hundred fifty-two isolates were studied by PFGE, and the most representative clones underwent MLST analysis (see text). Numbers in parentheses correspond to PT-A, -B, and -F, respectively. Not included are one PT-C isolate and one PT-D isolate found in 2000 and one PT-E isolate from 2002.