Activity of doripenem and comparator -lactams against US clinical isolates of Streptococcus pneumoniae with defined mutations in the penicillin-binding domains of pbp1a, pbp2b and pbp2x

Sir, Doripenem, a parenteral carbapenem, was recently approved in the USA for the treatment of complicated intraabdominal infections (cIAIs) and complicated urinary tract infections (cUTIs) including pyelonephritis. In Europe, a marketing authorization application has been filed for the treatment of cIAIs, cUTIs and nosocomial pneumonia. Doripenem has a broad spectrum of activity against clinically important pathogens including Enterobacteriaceae, Gram-negative non-fermenters, anaerobes and many Gram-positive cocci such as methicillin-susceptible Staphylococcus spp., group A streptococci and pneumococci.1 β-Lactam resistance in Streptococcus pneumoniae is caused by mutations in the penicillin-binding domains of one or more of its six penicillin-binding proteins (PBPs) resulting from mosaic genes or point mutations.2,3 Altered PBP1a, PBP2b and PBP2x are most important for β-lactam resistance in clinical isolates.2,3 This study reports the activity of doripenem against 30 S. pneumoniae US clinical isolates. β-Lactam MICs were determined using panels from Trek Diagnostic Systems (Cleveland, OH, USA) using CLSI recommendations.4 Breakpoints (non-meningitis) for all drugs except doripenem were those approved by CLSI.5 Currently, there are no approved doripenem breakpoints for S. pneumoniae. PBP gene sequencing and competition assays were performed as described previously.6 Seven penicillin-susceptible isolates had β-lactam MICs ≤0.03 mg/L (genotype wild-type) (Table 1). These isolates had no mutations in the penicillin-binding motifs of pbp1a, pbp2b and pbp2x. Of the eight penicillin-intermediate isolates (genotypes 1 and 2), three had penicillin MICs of 0.12 mg/L and MICs for the other β-lactams were ≤0.12 mg/L (genotypes 1A and 1B) (Table 1). They had a T446A substitution in the PBP2b SSNT motif; one isolate (genotype 1B) also had PBP2x substitutions (T338A and R384G). Five penicillin-intermediate isolates with penicillin MICs of 1 mg/L additionally had PBP2x substitutions of I371T and L546V (genotype 2, Table 1). The carbapenem MICs were the lowest of the agents tested, with imipenem having 2- and 4-fold lower MICs than doripenem and meropenem, respectively (genotype 2A). One isolate (genotype 2B) had substitutions of T371S in PBP1a and D623G in PBP2b. These substitutions were associated with 2–4-fold increases in carbapenem MICs (Table 1). Doripenem and comparator β-lactam MICs and PBP affinities in S. pneumoniae clinical isolates with defined PBP mutations aGT, genotype; wt, wild-type. bMIC range of each antimicrobial agent was determined for every genotype. The DOR, IPM, MEM and CRO IC50 (concentration of β-lactam that prevented 50% binding of Bocillin FL to PBP1a, PBP2x or PBP2b) was determined for one isolate each with wild-type genotype and 5C genotype. ND, not determined. cDOR, doripenem; IPM, imipenem; MEM, meropenem; CRO, ceftriaxone; AMC, amoxicillin/clavulanic acid; PEN, penicillin. dNo substitutions in the 557KTG motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. eNo substitutions in the 393AHSSNV motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. fNo substitutions in the 386SVVK motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. Doripenem and comparator β-lactam MICs and PBP affinities in S. pneumoniae clinical isolates with defined PBP mutations aGT, genotype; wt, wild-type. bMIC range of each antimicrobial agent was determined for every genotype. The DOR, IPM, MEM and CRO IC50 (concentration of β-lactam that prevented 50% binding of Bocillin FL to PBP1a, PBP2x or PBP2b) was determined for one isolate each with wild-type genotype and 5C genotype. ND, not determined. cDOR, doripenem; IPM, imipenem; MEM, meropenem; CRO, ceftriaxone; AMC, amoxicillin/clavulanic acid; PEN, penicillin. dNo substitutions in the 557KTG motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. eNo substitutions in the 393AHSSNV motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. fNo substitutions in the 386SVVK motif. Dash ‘-’ indicates no change from wild-type, and bold amino acids are different from wild-type. Among the penicillin-resistant isolates (genotypes 3–5), seven had penicillin MICs of 2–4 mg/L and were non-susceptible to imipenem and meropenem, three were non-susceptible to ceftriaxone and one was non-susceptible to amoxicillin/clavulanic acid; doripenem MICs were 0.5–1 mg/L (Table 1). These seven isolates (genotype 3) had PBP1a substitutions of T371A and P432T in addition to the PBP2x and PBP2b changes mentioned previously (Table 1). Two isolates with genotype 4 had additional substitutions of A619G and D623G in or near the KTGTA motif of PBP2b and a substitution of T371S in the STMK motif of PBP1a (Table 1). These changes corresponded with increased penicillin MICs (8 mg/L) and resistance to amoxicillin/clavulanic acid, imipenem and meropenem (Table 1). Doripenem MICs were 1–2 mg/L. Six isolates with genotype 5 had the same PBP substitutions as genotype 4 isolates, with the addition of the M339F change in PBP2x; in four isolates (genotypes 5A, 5B and 5C), there was also a Y595F substitution (Table 1). These changes were associated with a 2–16-fold increase in ceftriaxone MICs leading to resistance, whereas the MICs of the other drugs were within one doubling dilution when compared with genotype 4 isolates (Table 1). PBP binding studies with a wild-type isolate indicated that the carbapenems had good affinity for all six PBPs with IC50s ≤0.06 mg/L (Table 1). Ceftriaxone bound tightly to all PBPs (IC50s ≤ 0.1 mg/L) except PBP2b (Table 1) as expected, as cephalosporins do not use PBP2b as a primary target.2 In a genotype 5C isolate, the carbapenem binding affinities for all PBPs were reduced: PBP2b and PBP2x had the highest increase in IC50s (2–4 and 3.7–8 mg/L, respectively) (Table 1). The ceftriaxone IC50 for PBP2x was increased at least 200-fold when compared with strain 8865 (Table 1). β-Lactam MIC increases correlated with increases in the number of PBP1a, 2x and 2b substitutions. In this study, the PBP1a, 2b and 2x substitutions found in, and adjacent to, the penicillin-binding motifs SXXK, SXN and KT/SG were similar to substitutions reported by others.2,3 Carbapenems, like penicillins, are thought to have PBP2b as their primary target.3 In a study examining clinical isolates from Japan, a T624G PBP2b substitution was associated with carbapenem resistance.3 No isolates in our study had this mutation; however, a substitution at the adjacent amino acid (D623G) was found in seven of the eight genotype 4 and 5 isolates, all of which had elevated β-lactam MICs. This substitution was reported in a β-lactam-resistant S. pneumoniae recombinant that resulted from transformation with DNA from a β-lactam-resistant Streptococcus mitis strain.2 In summary, doripenem and imipenem (MIC90 1 mg/L) were 2-fold more active than meropenem (MIC90 2 mg/L) against the pneumococcal isolates in this study, including ceftriaxone-resistant isolates. The carbapenems had high affinity for all PBPs in a penicillin-susceptible isolate but had reduced binding affinity, particularly to PBP2x and PBP2b, in a penicillin-resistant isolate. This work was supported by Johnson & Johnson Pharmaceutical Research and Development, L.L.C. This work was presented in part at the Forty-sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, 2006 (Abstract C1-38). T. A. D., W. S., K. B. and R. K. F. own shares in Johnson & Johnson and are employees of Johnson & Johnson Pharmaceutical Research and Development, L.L.C.