The Ambler class D β-lactamase OXA-23–like subgroup is among the most common acquired mechanisms contributing to carbapenem resistance in Acinetobacter baumannii in the United States and globally (1). Detection of bla_OXA-23 has been infrequently reported outside of the genus Acinetobacter; bla_OXA-23–positive isolates of Proteus mirabilis are increasingly reported in Europe, particularly in France (2–7). Through routine testing conducted by the Centers for Disease Control and Prevention (CDC) Antimicrobial Resistance Laboratory Network (AR Lab Network), we detected a P. mirabilis clinical isolate from Massachusetts, USA, harboring bla_OXA-23. This activity was reviewed and approved by CDC as exempt human subjects research (see 45 C.F.R. part 46.104).

P. mirabilis was isolated in May 2024 from a urine culture of a patient receiving outpatient care in Massachusetts. In the 12 months before specimen collection, the patient had no reported history of international travel, domestic inpatient healthcare stays, or invasive procedures. As part of the AR Lab Network, clinical laboratories submit clinical isolates of carbapenem-resistant Enterobacterales to public health laboratories for characterization, as previously described (1) (Appendix).

We performed supplementary antimicrobial susceptibility testing at CDC by reference broth microdilution according to Clinical Laboratory Standards Institute (CLSI) guidelines as previously described (8); we interpreted results using CLSI breakpoints (9). The isolate was susceptible to both ertapenem and meropenem by broth microdilution and disk diffusion (Table), consistent with the characteristically weak carbapenemase activity of OXA-23 (2). The isolate was susceptible to all other β-lactams except ampicillin and doripenem (and also imipenem, to which Proteus spp. commonly have reduced susceptibility) (9) and all tested cephalosporins apart from cefazolin (Table). Of note, the isolate showed susceptibility to piperacillin/tazobactam (Table), which is less frequently observed for bla_OXA-23–positive isolates (6). The isolate was positive for carbapenemase production but PCR-negative for the 5 carbapenemase genes targeted by the AR Lab Network.

Figure

Figure. Genomic comparisons of bla_OXA-23–positive Proteus mirabilis detected in Massachusetts, USA, in 2024 to clinical isolates identified in previous studies. A) Genomic context of bla...

We performed whole-genome sequencing (WGS) and data analysis; we submitted WGS data for the strain, 2024DK-00154, to the National Center for Biotechnology Information BioSample database (https://www.ncbi.nlm.nih.gov/biosample; accession no. SAMN41612253) (Appendix). We obtained a closed genome of 3.93 Mb that was not found to harbor any plasmids. The β-lactamase gene bla_OXA-23 and 9 other AR genes were detected in this isolate (tetJ, aadA1, dfrA1, sat2, sul2, aph(6)-Id, aph(3”)-Ib, aac(3)-IIe, aph(3′)-Ia). The bla_OXA-23 gene is located inside a Tn2008-like transposon (Tn6704) within a Tn6703-like genomic island, closely resembling the bla_OXA-23 region of strains from Switzerland (Pm1 and Pm5) (7) and France (VAC) (4) (Figure, panel A).

We inferred a phylogenetic tree of 2024DK-00154 and 12 other bla_OXA-23-positive P. mirabilis strains on the basis of the alignment of the core genome (≈3.23 Mb; 81% of the average genome size) (Figure, panel B). The 2024DK-00154 strain clustered with clinical isolates collected in Europe during 1996–2023, including an OXA-23–producing lineage circulating in France and Belgium since 1996 (2,4) and closely related strains detected in Germany (6) and Switzerland (7) (Figure, panel B). Within that cluster, all isolates belonged to sequence type (ST) 142 and differed by 4–200 core-genome single-nucleotide variants (SNVs); 2024DK-00154 is most similar to a strain isolated in Germany (Carb-21), differing by only 27 SNVs across ≈80% of the genome. In contrast, 2024DK-00154 differs from non-ST142 isolates from Singapore and France in which bla_OXA-23 is plasmidborne (4,10) by 17,321–18,347 SNVs.

Those results suggest that the P. mirabilis isolate identified in Massachusetts is closely related to contemporary bla_OXA-23–positive ST142 P. mirabilis strains circulating in Europe, consistent with international dissemination. In France, OXA-23–producing P. mirabilis strains (that cluster with the Massachusetts isolate) have been recovered from epidemiologically unrelated patients, suggesting community spread (3). Community spread has not been detected in the United States but is possible. Indeed, the Massachusetts case-patient had no reported international travel, suggesting a local infection source.

Testing for carbapenemase genes in the AR Lab Network is initiated on the basis of the detection of carbapenem resistance at the clinical laboratory. Thus, bla_OXA-23–positive P. mirabilis isolates, which are characteristically carbapenem-susceptible, would not be targeted for mechanism testing, and the presence of bla_OXA-23 would go undetected. Consequently, appropriate infection control measures that would be taken for patients identified as harboring carbapenemase-producing organisms would not necessarily be taken; P. mirabilis could serve as an unmonitored reservoir enabling ongoing transmission of bla_OXA-23 to other species, including other Enterobacterales species. The inability to detect reservoirs of carbapenemase genes, including environmental sources, healthcare workers, or novel gene-organism combinations, could pose challenges for identifying and addressing sources of transmission.

In addition to displaying susceptibility to carbapenems, bla_OXA-23–harboring P. mirabilis might also escape detection during routine laboratory testing for carbapenemases because of poor sensitivity of phenotypic carbapenemase production tests and the absence of that target from most commercial PCR and immunochromatographic tests (6). A diagnostic algorithm based on antimicrobial susceptibility profiles and phenotypic assays was recently developed to improve detection of OXA-23 and other carbapenemases whose frequency might be underestimated in Proteus spp. (3,4,6). Our findings suggest the need for continued vigilance in detecting bla_OXA-23 in genera beyond Acinetobacter to prevent further spread between taxa.

Top