In vitro activity of cefiderocol against Pseudomonas aeruginosa demonstrating evolved resistance to novel β-lactam/β-lactamase inhibitors

Abstract Background Cefiderocol demonstrates excellent activity against MDR Pseudomonas aeruginosa; however, the activity against isolates from patients previously treated with β-lactam agents is unknown. We aimed to determine the activity of cefiderocol against P. aeruginosa collected before and after treatment with traditional β-lactams and new β-lactam/β-lactamase inhibitors. Methods Cefiderocol MICs were determined in triplicate in iron-depleted cation-adjusted Mueller–Hinton broth and compared with β-lactam MICs tested by standard methods. All isolates underwent WGS analysis to identify mutations associated with resistance. Results One hundred and seventy-eight P. aeruginosa isolates were evaluated; 48% (86/178) were non-susceptible to ceftazidime/avibactam, ceftolozane/tazobactam and/or imipenem/relebactam. The cefiderocol MIC50 and MIC90 were 0.12 and 1 mg/L, respectively. Median cefiderocol MICs did not vary against isolates classified as MDR, XDR, or those non-susceptible to ceftazidime/avibactam, ceftolozane/tazobactam and/or imipenem/relebactam when compared with non-MDR isolates. Against isolates collected from patients previously treated with ceftolozane/tazobactam, cefiderocol MICs were increased 4-fold compared with baseline. Cross-resistance to cefiderocol was identified in 21% (3/14) of patients who developed treatment-emergent resistance to ceftolozane/tazobactam. Overall, 6% (11/178) of isolates demonstrated cefiderocol MICs ≥2 mg/L, which were disproportionately collected from patients previously treated with ceftolozane/tazobactam (73%; 8/11). Isolates with reduced cefiderocol susceptibility harboured mutations in ampC, tonB-dependent receptors, the response regulator pirR and ftsI. Conclusions Cefiderocol demonstrates excellent in vitro activity against P. aeruginosa isolates exposed to other novel β-lactam agents; however, some exceptions were identified. Cross-resistance between cefiderocol and ceftolozane/tazobactam was evident, but not with ceftazidime/avibactam or imipenem/relebactam. Reduced cefiderocol susceptibility was mediated by mutations in ampC and tonB-dependent receptors.


Introduction
3][4] Cefiderocol is unique when compared with the newest generation of β-lactam/β-lactamase inhibitor agents given that it uses iron transport systems to gain entry into the bacterial cell.The agent demonstrates stability against most known mechanisms of antibiotic resistance, including MBLs. 5 In surveillance studies, cefiderocol demonstrates high rates of in vitro activity against MDR P. aeruginosa, including against isolates that are resistant to ceftazidime/avibactam, ceftolozane/tazobactam and/or imipenem/relebactam. [6][7][8][9][10] Although encouraging, isolates included in these studies were not collected from patients previously treated with novel β-lactams, and differential resistance rates may be due to the presence of MBLs.In fact, underlying resistance mechanisms are often not well characterized in surveillance studies, particularly for P. aeruginosa where β-lactam resistance is mediated by multiple adaptive mechanisms.
Given the potential utility of cefiderocol for treatment of refractory MDR P. aeruginosa infections, we sought to determine its in vitro activity against genetically diverse clinical P. aeruginosa isolates and assess the impact of prior β-lactam treatment.We hypothesized that prior treatment and resistance to traditional β-lactam agents would not impact cefiderocol activity; however, the development of resistance to novel β-lactam/β-lactamase inhibitors would lead to reduced susceptibility.Thus, we selected P. aeruginosa isolates from patients at our centre not previously treated with β-lactams (n = 31), isolates from patients previously treated with traditional β-lactams (n = 69), and serial isolates from patients before and after treatment-emergent resistance to novel β-lactam/β-lactamase inhibitors (n = 78).We used WGS to identify underlying mechanisms of resistance and evaluate cross-resistance between cefiderocol and other agents.

Materials and methods
One hundred and seventy-eight P. aeruginosa isolates were selected for investigation, including isolates from an ongoing prospective study of healthcare-associated infections. 11MDR and XDR isolates were defined by consensus criteria. 12Prior antibiotic exposure was defined as receipt of β-lactam treatment for ≥3 days within 90 days prior to the isolate collection date.For patients treated with novel β-lactam/β-lactamase inhibitor agents, serial isolates were included to compare baseline and post-exposure isolates from the same patient.This resulted in the inclusion of 78 isolates collected from patients before and after treatment with ceftolozane/tazobactam (n = 14 patients, 36 isolates), ceftazidime/ avibactam (n = 4 patients, 13 isolates) and imipenem/relebactam (n = 5 patients, 29 isolates).None of the patients had received prior treatment with cefiderocol at the time of isolate collection.Clinical characteristics of patients treated with ceftolozane/tazobactam and imipenem/ relebactam have been reported previously. 4,13ICs were determined in triplicate by broth microdilution (BMD) methods according to CLSI guidelines to identify the modal MIC. 14,15mipenem (0.03-32 mg/L) was tested with and without relebactam (4 mg/L); meropenem (0.06-64 mg/L) was tested with and without vaborbactam (8 mg/L).Ceftolozane (0.06-64 mg/L) and piperacillin (0.05-512 mg/L) were both tested with 4 mg/L of tazobactam.Ceftazidime (0.25-256 mg/L) was tested with and without avibactam (4 mg/L).Aztreonam (0.12-128 mg/L) and cefepime (0.25-256 mg/L) were also tested for completeness.Cefiderocol panels (range 0.03-32 mg/mL) were provided by IHMA (Schaumberg, IL, USA) and made onsite using Chelex ® -treated, iron-depleted, cation-adjusted Mueller-Hinton broth. 15uality control (QC) strains P. aeruginosa ATCC 27853, Klebsiella pneumoniae ATCC 700603, Escherichia coli ATCC 25922 and K. pneumoniae BAA-1705 were used throughout, and results were only included when QC strains were within CLSI reference ranges.All MICs were interpreted according to CLSI interpretive criteria. 15GS was performed and analysed as previously described. 4,11,13riefly, DNA was sequenced on the Illumina platform, assembled using SPAdes 3.14 and annotated with Prokka. 16,17SNPs were called using SAMtools, and MLST was performed with MLST v2.19 (https://github.com/tseemann/mlst). 18The presence or absence of β-lactamase genes was confirmed using ResFinder. 19Following a literature search, 19 genes of interest were chosen a priori for analysis (Table S1, available as Supplementary data at JAC-AMR Online).Protein sequences were compared with P. aeruginosa PAO1.Genomes are available under NCBI BioProjects PRJNA475751, PRJNA715186, PRJNA782612 and PRJNA872132.
Categorical and continuous variables were analysed using a chisquared (or Fisher's exact) test and Mann-Whitney U test, respectively.A two-tailed P value ≤0.05 was considered statistically significant.
To further explore the impact of treatment-emergent resistance to these novel agents, cefiderocol susceptibility was evaluated against 36 isolates (15 baseline, 21 post-exposure) from 14 patients that developed treatment-emergent resistance to ceftolozane/tazobactam. 4 Baseline isolates from 12 patients belonged to unique STs; the final two isolates each had a unique allele profile that was not identifiable within the MLST database.Five PDC variants were present at baseline; PDC-5 was most Shields et al.  common (n = 6).All isolates carried bla OXA-50-like ; however, no MBL or other serine β-lactamase genes known to confer ceftolozane/ tazobactam resistance were found. 21,22Following ceftolozane/ tazobactam treatment [median (range) duration = 19 (3-52) days], median ceftolozane/tazobactam MICs increased from 2 mg/L at baseline to 64 mg/L after exposure (P < 0.0001).Comparable MICs changes were noted for ceftazidime/avibactam (from 4 to 64 mg/L, P < 0.0001; Table 1).Treatment-emergent mutations in PDC were identified in 79% (11/14) of patients. 4Overall, median cefiderocol MICs increased from 0.12 mg/L among baseline isolates to 0.5 mg/L among isolates collected after ceftolozane/tazobactam exposure (P = 0.1034; Table 1).Fourteen percent (3/21) of post-exposure isolates yielded cefiderocol MICs >4 mg/L; each produced undefined variant PDC and harboured mutations in pirR, a regulator of the tonB-dependent receptor (TBDR) pirA.Two of the three resistant isolates also contained mutations in ftsI, which encodes penicillin-binding protein 3 (PBP3) in P. aeruginosa.
Median cefiderocol MICs were increased from 0.12 to 0.25 mg/L against isolates collected after exposure to any one of ceftazidime/avibactam, ceftolozane/tazobactam or imipenem/relebactam (n = 47) compared with baseline isolates from the same patients (n = 31; P = 0.1442).
Eleven isolates from 10 unique patients demonstrated cefiderocol MICs above the MIC 90 (cefiderocol MIC ≥2 mg/L).To determine genetic factors associated with reduced cefiderocol susceptibility, we assessed mutations in relevant genes by comparing the protein sequence of each gene with the PAO1 sequence (Table S1). 9,23,24Mutations in ampC, AmpC regulators (ampD, ampR), ftsI or oprD were encountered in all 11 isolates.The inner membrane proteins exbB and exbD were each WT in 91% (10/11) of isolates (Table 2).Sixty-four percent (7/11) of isolates harboured ftsI mutations; however, the presence or absence of ftsI mutations alone did not impact cefiderocol MICs (median MIC = 0.25 mg/L for both groups).Two isolates contained major disruptions in either tonB2 or tonB3.Three additional TBDR systems were analysed, including fecA (with fecI and fecR), piuA (or piuD, with piuC) and pirA (with pirR and pirS).Compared with the PAO1 P. aeruginosa reference sequence, mutations were identified in each TBDR system.Among serial isolates collected from the same patient, sequence changes were observed in piuD and pirR, but not fecA, fecI, fecR, piuA, piuC or pirS.In the pirA TBDR system that includes pirR and pirS, 54.5% (6/11) of isolates contained a single base pair insertion or deletion in a poly(dC) tract of pirR causing an early termination at amino acid position 201.Mutations in pirA were identified in all isolates and did not correlate with cefiderocol MICs.On the other hand, frameshift mutations in pirR were identified in 45% (5/11) isolates with reduced FDC susceptibility.Across isolates from patients treated with ceftolozane/tazobactam, ceftazidime/avibactam or imipenem/relebactam (n = 78 total), median cefiderocol MICs were 4 and 0.25 mg/L in isolates with or without a frameshift in pirR (P < 0.0001).

Discussion
Our data corroborate and extend prior surveillance studies describing the excellent in vitro activity of cefiderocol against P. aeruginosa clinical isolates.7][8][9] Against MDR and XDR isolates, the cefiderocol MIC 50 /MIC 90 were 0.06/0.25 and 0.25/1 mg/L, respectively, confirming that commonly encountered mechanisms of antibiotic resistance in P. aeruginosa do not significantly contribute to reduced cefiderocol susceptibility.To extend these findings, we evaluated the impact of prior treatment with β-lactam antibiotics on the activity of cefiderocol.Not surprisingly, we found that exposure and resistance to traditional β-lactams did not impact the in vitro activity of cefiderocol.Next, we tested P. aeruginosa isolates recovered from patients previously treated with ceftazidime/ avibactam, ceftolozane/tazobactam or imipenem/relebactam to represent cases where cefiderocol is likely to be considered as a treatment option.Against such isolates, the corresponding cefiderocol MIC 50 /MIC 90 were 0.06/0.25,0.25/4 and 0.25/2 mg/L, respectively.Median cefiderocol MICs were generally not impacted following exposure to any of these agents individually (Table 1); however, some exceptions were noted.Most importantly, median cefiderocol MICs were increased 4-fold following ceftolozane/ tazobactam exposure, and the emergence of cross-resistance between ceftolozane/tazobactam and cefiderocol was demonstrated in 21% (3/14) of patients.To put these findings into context, only 6% (11/178) of all isolates studied demonstrated cefiderocol MICs ≥2 mg/L, and 73% (8/11) of these isolates were collected from patients previously treated with ceftolozane/ tazobactam.
Our findings of potential cross-resistance between ceftolozane/tazobactam and cefiderocol are in agreement with prior reports. 23,25,26The underlying mechanisms of resistance appears to include mutations in ampC and TBDR genes, specifically pirR, a   piuD is an orthologue of piuA.

Shields et al.
regulator of TBDR gene piuA.8][29] During in vitro selection studies an L320P mutation in the R2 loop of AmpC was selected across multiple P. aeruginosa lineages after exposure to cefiderocol. 26We hypothesize that these conformational changes in PDC known to impact ceftolozane/tazobactam also affect the affinity for cefiderocol, an observation we previously demonstrated in clinical isolates of Enterobacter cloacae complex. 27,30In this regard, it is notable that isolates in the current study with the highest cefiderocol MICs (>4 mg/L) were universally resistant to ceftolozane/tazobactam and harboured mutations in ampC.Unlike other ceftolozane/tazobactam-resistant isolates, however, those with elevated cefiderocol MICs also encoded mutations in pirR, most commonly a single base-pair insertion or deletion in the same poly(dC) tract resulting in a frameshift and loss of function (Table 2, Table S3).Frameshift mutations in pirR were identified among isolates collected from patients treated with ceftolozane/tazobactam, and importantly not among surveillance isolates, or those collected from patients treated with ceftazidime/avibactam or imipenem/relebactam.PirR encodes the response regulator of the two-component PirRS system that regulates expression of TBDRs.The TBDR PiuA, and its orthologue PiuD, have been reported as the main transporters for cefiderocol as decreased expression of associated genes resulted in 128-fold increased cefiderocol MICs. 31,32We did not observe relevant or unique changes to piuA or piuD genes, which may be due to the fact that none of the patients in the present study were previously treated with cefiderocol.Moreover, we did not identify new mutations in piuC, which has been implicated in the development of cefiderocol resistance previously. 26Finally, it appears that decreased expression of TBDR PirA may influence cefiderocol MICs, although to a lesser extent by comparison, 31 which aligns with the data reported herein.Detection of cefiderocol non-susceptible P. aeruginosa following treatment with ceftolozane/tazobactam was first reported by Streling and colleagues 23 in a single case.In that prior report, a patient treated with ceftolozane/tazobactam for 14 days developed treatment-emergent ceftolozane/tazobactam resistance, and a P. aeruginosa subpopulation demonstrated an increased cefiderocol MIC from 2 mg/L at baseline to 8 mg/L following ceftolozane/tazobactam exposure.WGS identified disruptions in both piuD and pirR.To contrast these findings, a study of 32 paired isolates collected before and after ceftolozane/tazobactam exposure in 16 patients did not identify any mutations in TBDR genes; however, at least 4-fold increases in cefiderocol MIC among post-exposure isolates were noted in 25% of cases. 25hese cases were associated with new mutations in ampC, ampD and mexR.In our study, 57% (8/14) of ceftolozane/ tazobactam-treated cases were associated with at least a 4-fold increase in cefiderocol MICs following exposure, and 21% (3/14) resulted in cefiderocol resistance when defined by an MIC >4 mg/L.Resistance in these cases was associated with the presence of mutations in both ampC and TBDR genes.The cumulative data suggest that cross-resistance between ceftolozane/tazobactam and cefiderocol is relatively uncommon, but an important phenomenon that has now been documented to varying degrees by three independent studies.Mutations in ampC alone are associated with ∼4-fold increased cefiderocol MICs that remain within the susceptible range (MIC ≤4 mg/L) for most isolates.Additional mutations in TBDR are likely necessary to confer resistance to cefiderocol, and seemingly arise more commonly following treatment with ceftolozane/tazobactam than other novel β-lactam/β-lactamase inhibitor combinations.These data serve as a caution for use of cefiderocol following the emergence of ceftolozane/tazobactam resistance without first confirming susceptibility.
We did not identify cross-resistance or significantly reduced susceptibility to cefiderocol among patients treated with ceftazidime/avibactam or imipenem/relebactam, which was evaluated in fewer patients overall.Among serial isolates from five patients who developed resistance to imipenem/relebactam attributed to mutations in the mexAB-oprM or mexEF-oprN efflux operons, none demonstrated a cefiderocol MIC >2 mg/L or a >4-fold increased cefiderocol MIC relative to the baseline isolate from each patient.These findings were consistent among serial isolates from patients treated with ceftazidime/avibactam where post-exposure isolates from three patients harboured mutations in mexAB-oprM or its repressor gene mexR, and each had cefiderocol MICs ≤0.25 mg/L.Our data support the hypothesis that cefiderocol is not a major substrate for efflux in P. aeruginosa.In fact, against engineered strains either deficient in or overexpressing mexAB-oprM, cefiderocol MICs were only shifted 2-fold. 32hat said, paired clinical isolates from two patients treated with ceftolozane/tazobactam that demonstrated ≥4-fold cefiderocol MICs only showed mutations in mexR (A66V or L57D), but not other genes investigated. 25We also identified mutations at the L57 position in mexR (L57P and L57Q), and consistent with the previous report, cefiderocol MICs with mutations in mexR ranged from 0.25 to 2 mg/L.Other mutations (D89E) in mexR have been associated with increased cefiderocol MICs in a case report. 33rom these data we propose that efflux is not a primary contributor to reduced cefiderocol susceptibility, but overexpression of efflux pumps may play a minor role in isolates that also overexpress ampC.The same is likely true of mutations in ftsI, which encodes PBP3, the primary inhibition target for cefiderocol.We identified ftsI mutations in both cefiderocol-susceptible and -resistant P. aeruginosa isolates without clear associations.We previously reported that ftsI mutations are complementary to, but not primarily responsible for, ceftolozane/tazobactam resistance, 4 and were identified in one patient with treatment-emergent resistance to ceftazidime/avibactam in the present report.
Our data add to a growing number of reports that have identified mechanisms of cefiderocol reduced susceptibility in P. aeruginosa clinical isolates.Reported mechanisms broadly include mutations in iron transport systems, two component response regulators, TBDRs, ampC and efflux operons. 7,23,25,31,33,34otably, few reports have described the emergence of cefiderocol resistance following treatment with the agent.In the CREDIBLE-CR study, 15% of isolates collected from patients treated with cefiderocol demonstrated a ≥4-fold MIC increase, including three cases involving P. aeruginosa. 9WGS of posttreatment P. aeruginosa isolates did not identify relevant mutations in two cases (notably pirR mutations were not investigated).In the third case, however, a four amino acid deletion in PDC was identified, and subsequently associated with an 8-fold cefiderocol MIC increase in isogenic strains.The data further Cefiderocol MICs against BL/BLI-resistant Pseudomonas support the role of ampC mutations in reduced cefiderocol susceptibility following treatment with cephalosporins. 27,30It remains to be seen how readily TBDR gene mutations are selected for following treatment with cefiderocol.In two previously reported cases, cefiderocol resistance emerged following treatment of XDR P. aeruginosa infections; however, molecular analysis of the resistant isolates was not conducted. 35,36n summary, we offer new and important insights for the in vitro activity of cefiderocol against MDR and XDR P. aeruginosa isolates for which cefiderocol treatment is likely to be considered.Indeed, nearly half of isolates included in this study were resistant to ceftolozane/tazobactam, ceftazidime/avibactam and/or imipenem/relebactam, including serial isolates from patients treated with these agents.The most noteworthy finding resulting from our analysis is the potential for reduced cefiderocol activity in the setting of treatment-emergent resistance to ceftolozane/tazobactam.This is particularly problematic given crossresistance between ceftolozane/tazobactam and ceftazidime/ avibactam mediated by amino acid substitutions in the catalytic centre of PDC that result in structural changes. 4,25Indeed, structural modifications to PDC appear to have a negative impact on all cephalosporin-based antipseudomonal agents, but fortunately are associated with collateral sensitivity to imipenem. 4,37Thus, treatment with imipenem/relebactam may be the most reasonable option in the setting of treatment-emergent resistance to ceftolozane/tazobactam. Susceptibility to both cefiderocol and imipenem/relebactam should be confirmed prior to treatment initiation given that clinical experience with either agent for treatment of MDR and XDR P. aeruginosa remains limited.It is also important to highlight that treatment-emergent resistance to ceftazidime/avibactam and imipenem/relebactam did not significantly attenuate the activity of cefiderocol.Mechanisms of resistance to these agents include increased expression of ampC and efflux pumps that minimally impact cefiderocol in vitro activity.Overall, this is consistent with the broader potency of cefiderocol against P. aeruginosa.Finally, it should be noted that further validation is needed to confirm mechanisms of reduced cefiderocol activity, and in particular the association of adaptive TBDR gene mutations following treatment with ceftolozane/ tazobactam.Nevertheless, the in vitro activity of cefiderocol against this challenging set of P. aeruginosa isolates supports routine cefiderocol testing for clinical isolates where antibiotic options are limited.

Table 2 .
Characterization of antibiotic MICs and relevant antibiotic resistance genes for isolates with cefiderocol MICs ≥2 mg/L