AMRrounds: What goes around comes back around—recurrent KPC-producing Salmonella bacteraemia

Our patient is an 81-year-old male with transfusion-dependent VEXAS (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) syndrome, smouldering IgM multiple myeloma, prostate adenocarcinoma and bioprosthetic transcatheter aortic valve replacement, receiving long-term combination daptomycin and ceftaroline for aortic valve endocarditis with perivalvular abscess due to methicillin-resistant Staphylococcus epidermidis. The patient was initially admitted to the ICU from home with fever and septic shock. Empirical meropenem 500 mg IV q8 h was started and daptomycin and ceftaroline were continued. Blood cultures revealed Salmonella spp. (Figure 1) and the isolate was sent to the state public health lab for additional testing. Meropenem was changed to trimethoprim/sulfamethoxazole 1 DS tablet q8 h based on susceptibilities with an anticipated 6 week course due to a suspected intravascular source. At the time of discharge, limited susceptibilities were known, as seen in Figure 1. Susceptibilities with CLSI and EUCAST interpretations for the Salmonella group (Salmonella Enteritidis per state laboratory) isolate from blood cultures on initial hospital admission. aCLSI.2 bEUCAST.9 cAzithromycin testing is only recommended for S. Typhi. Performed per clinician request. dCLSI notes that the susceptible breakpoint of MIC ≤16 mg/L is based on MIC distribution data and limited clinical data. eEUCAST notes that isolates with MIC >16 mg/L are likely to have resistance mechanisms. fCLSI recommends I^ for zones 20–29 mm; however, interpretation was reported as R on the basis of institutional guidelines due to concern for clinical failure or delayed response to quinolone treatment when a quinolone is selected for non-susceptible isolates. gEUCAST does not recommend disc diffusion with 5 µg ciprofloxacin disc. Only MIC tests or inferred susceptibility from 5 µg pefloxacin disc are recommended. Pefloxacin discs are not available in the USA. hDoxycycline is considered unsuitable. iAt the time of initial hospital discharge, meropenem was reported as I based on Kirby–Bauer disc diffusion and CLSI interpretation. This was changed to R based on later Xpert® Carba-R PCR testing, performed after the patient’s second hospitalization. jPer EUCAST interpretation, this zone (<28 mm) is within the ATU and the isolate should be screened for carbapenemase. kMeasured per CLSI, disregarding slight (≤20% of lawn) growth and measuring the more obvious zone of inhibition. This effect is due to antagonists, such as thymine and thymidine, in media. Two days after hospital discharge, the patient was found to be unresponsive and readmitted from an outside facility. Blood cultures at the outside facility were positive for Salmonella spp. with an unknown rapid test positive for Klebsiella pneumoniae carbapenemase (KPC). Blood cultures from our facility were negative after 5 days. Given the reported KPC from the outside hospital and meropenem non-susceptibility, our laboratory performed an Xpert® Carba-R (Cepheid) on our isolate from his initial hospitalization, and it was positive for KPC. What is the most likely explanation for this resistance profile and recommendation for treatment? Salmonella spp. are members of the Enterobacterales family and are facultative anaerobic, flagellated Gram-negative rods.1 Salmonella spp. are a common cause of infectious diarrhoea in humans. Transmission may occur via consumption of contaminated animal products or produce, person-to-person contact via faecal–oral route or from contact with pets. Salmonella spp. do not carry intrinsic resistance.2 However, aminoglycosides, first and second-generation cephalosporins and cephamycins may not be clinically active due to production of beta lactamases or the intracellular nature of Salmonella spp. lifecycle. Selected acquired mechanisms of resistance are described in Table 1. In our clinical microbiology laboratory, identification was performed using MALDI-TOF on colonies grown from blood culture. Susceptibility was performed by disc diffusion for most antimicrobials and azithromycin testing was performed by Etest following clinician request. CSLI recommends a tiered approach to testing for Salmonella spp., with standard testing for extra-intestinal isolates performed for ampicillin, ceftriaxone, ciprofloxacin and trimethoprim/sulfamethoxazole only.2 Selected acquired resistance mechanisms reported in Salmonella spp. Salmonella does not carry intrinsic resistance, but aminoglycosides, first- and second-generation cephalosporins and cephamycins are not clinically effective.2 aFor more details on substrates and inhibitors, please refer to https://pmc.ncbi.nlm.nih.gov/articles/instance/10315565/bin/bmb-56-6-326-supple.pdf (supplemental table from Jang14). Selected acquired resistance mechanisms reported in Salmonella spp. Salmonella does not carry intrinsic resistance, but aminoglycosides, first- and second-generation cephalosporins and cephamycins are not clinically effective.2 aFor more details on substrates and inhibitors, please refer to https://pmc.ncbi.nlm.nih.gov/articles/instance/10315565/bin/bmb-56-6-326-supple.pdf (supplemental table from Jang14). As we consider the resistance patterns observed in our particular isolate of Salmonella, it is useful to note common phenotypes such as Newport strain REPJJP0, which is a multi-drug-resistant serotype originally reported in 2015 and frequently associated with major outbreaks.3 Newport is considered genetically diverse, and resistance monitoring performed by the CDC associates it with resistance to trimethoprim/sulfamethoxazole (86%), tetracycline (86%), azithromycin (86%), ciprofloxacin (70%) and ampicillin (68%); notably, all isolates tested were susceptible to meropenem.3 Our isolate’s phenotype differs from the Newport strain in its susceptibility to trimethoprim/sulfamethoxazole and its resistance to ceftriaxone and meropenem. This isolate was not linked to a known outbreak, and we do not have serovar or genetic typing to allow closer comparison with outbreak strains. The β-lactam resistance observed in this isolate was due to production of an Ambler class A carbapenemase, KPC. Carbapenemase production in Enterobacterales is a major public health concern, although carbapenem susceptibility testing is not routinely recommended by CLSI.2,4 Testing for carbapenemase may be useful when one or more of the carbapenems tests as non-susceptible. Carbapenemase production in Salmonella spp. is reported sporadically and is not as prevalent as it is in other members of the family Enterobacterales; however, all three carbapenemase classes, A, B and D, have been identified rarely in Salmonella spp.5–8 Interestingly, the original disc diffusion categorical interpretation of meropenem was intermediate as defined by CLSI (20–22 mm)2 while EUCAST recommendations would have designated area of technical uncertainty and prompted screening for carbapenemase production.9 Our local laboratory followed CLSI guidelines and did not originally pursue carbapenemase screening because other agents were available for treatment; further testing was deferred to the state public health laboratory. We later performed the Xpert® Carba-R (Cepheid) molecular test, and the state public health lab identified KPC on the index isolate using the rapid, lateral flow NG-Test® CARBA 5 assay (Hardy Diagnostics). On the basis of the Carba-R result, the meropenem interpretation was changed to resistant to discourage use of a carbapenem. In line with KPC activity, both B-lactam/B-lactamase inhibitor combinations were susceptible. Therapy was changed from meropenem 1 g IV q8 h to meropenem/vaborbactam 4 g IV q8 h on hospital Day 2 of the second admission, which the patient continued for 42 days after discharge. Our isolate also demonstrated reduced susceptibility to ciprofloxacin. While the disc diffusion zone of 29 mm corresponds to an intermediate category (21–30 mm designated I^) per CLSI,2 we reported the isolate resistant per institutional guidelines. Isolates that do not show full susceptibility to quinolones are associated with delayed response or clinical failure when quinolones are selected for therapy.2 Among other mechanisms, one or multiple mutations in the DNA gyrase genes gyrA or gyrB in the quinolone resistance determining region (QRDR) confers resistance or reduced susceptibility to ciprofloxacin.10 It is likely that our strain acquired mutations in one or more of these genes to confer quinolone resistance. Our isolate tested resistant to doxycycline. Tetracycline resistance is also relatively common in Salmonella spp., and can occur due to changes in efflux, antibiotic destruction, binding site protection or modification of the tetracycline binding site on the 16S rRNA. Of eight tet genes associated with increased efflux of tetracyclines, tet(A) is the most common in Salmonella and Escherichia coli.11,12 Tetracycline, like carbapenems, is not recommended by CLSI for routine testing for Salmonella isolates, but may be tested based on clinician request when other agents are not suitable.2 Last, this isolate had a low MIC to azithromycin that would be considered susceptible based on Salmonella enterica serovar Typhi breakpoints. While our laboratory does not routinely test azithromycin, this test was performed per clinician request due to the extensive drug resistance shown. CLSI notes that the susceptible breakpoint of MIC ≤16 mg/L is based on MIC distribution data and limited clinical data.2 EUCAST notes that isolates with MIC >16 mg/L are likely to have resistance mechanisms, such as efflux pumps or macrolide-inactivating enzymes.9,13 Salmonella may harbour diverse mechanisms of resistance that compromise key therapeutic drug classes. The resistant strain encountered in this case was assessed using the standardized framework for AMRrounds. In this case, clinical laboratory testing methods and standard setting organization guidance impacted the time to detection of resistance and the treatment options selected. Elevated MIC or smaller disc diffusion zone to a carbapenem should be considered a potential marker of carbapenemase production and should prompt additional testing at local clinical laboratories. Our patient with KPC-producing Salmonella bacteraemia received 6 weeks of meropenem/vaborbactam before relapsing a second time and being retreated with an 8-week course of meropenem/vaborbactam. He remains on chronic suppressive trimethoprim/sulfamethoxazole and continues to follow in our clinic with no further relapses. This work was unfunded. None to declare. All authors, study conception and design; A.H.M. and C.N.D., data collection and preparation of the first manuscript draft; all authors, manuscript preparation—critical revision and contribution. Patient informed consent was obtained.