ABSTRACT
Cefmetazole is active against extended-spectrum β-lactamase-producing Escherichia coli (ESBLEC) and is a potential candidate for carbapenem-sparing therapy. This multicenter, observational study included patients hospitalized for invasive urinary tract infection due to ESBLEC between March 2020 and November 2021 at 10 facilities in Japan, for whom either cefmetazole or meropenem was initiated as a definitive therapy within 96 h of culture collection and continued for at least 3 d. Outcomes included clinical and microbiological effectiveness, recurrence within 28 d, and all-cause mortality (14 d, 30 d, in-hospital). Outcomes were adjusted for the inverse probability of propensity scores for receiving cefmetazole or meropenem. Eighty-one and forty-six patients were included in the cefmetazole and meropenem groups, respectively. Bacteremia accounted for 43% of the cefmetazole group, and 59% of the meropenem group. The crude clinical effectiveness, 14 d, 30 d, and in-hospital mortality for patients in the cefmetazole and meropenem groups were 96.1% vs 90.9%, 0% vs 2.3%, 0% vs 12.5%, and 2.6% vs 13.3%, respectively. After propensity score adjustment, clinical effectiveness, the risk of in-hospital mortality, and the risk of recurrence were similar between the two groups (P = 0.54, P = 0.10, and P = 0.79, respectively). In all cases with available data (cefmetazole : n = 61, meropenem : n = 22), both drugs were microbiologically effective. In all isolates, bla CTX-M was detected as the extended-spectrum β-lactamase gene. The predominant CTX-M subtype was CTX-M-27 (47.6%). Cefmetazole showed clinical and bacteriological effectiveness comparable to meropenem against invasive urinary tract infection due to ESBLECs.
KEYWORDS: antimicrobial resistance, carbapenem, E. coli, urinary tract infection, cephamycin
INTRODUCTION
Third-generation cephalosporin-resistant (3GCR) Escherichia coli has been increasing worldwide. The reported median 3GCR rate of bacteremia due to E. coli, which is one of the two sustainable development goals for antimicrobial resistance indicators, was 36.6% [interquartile range (IQR), 17.5–58.3] according to the recent global surveillance report (1). Based on data from 2,167 medical institutions in Japan, 3GCR E. coli was isolated from 3.7% of hospitalized patients (i.e., approximately 100,000 patients per year). The 3GCR rate of E. coli has continually been increasing (26.8% in 2017 and 28.3% in 2020) (2).
Extended-spectrum β-lactamase (ESBL) production is the main mechanism by which E. coli acquires resistance to third-generation cephalosporins (1). Carbapenems are the first-line treatment option for infections due to ESBL-producing E. coli (ESBLEC). However, increased carbapenem usage may exert selective pressure on the indigenous flora, leading to disadvantages including an increase in carbapenem-resistant bacteria (3). Therefore, there is a need for effective and targeted carbapenem-sparing therapy for ESBLEC infection. The most promising carbapenem-sparing therapy was piperacillin-tazobactam; however, a recent international randomized clinical trial was not able to demonstrate its noninferiority to meropenem (MEM) (4).
Cefmetazole (CMZ), a semisynthetic cephamycin antibiotic, is stable against hydrolysis by ESBLs and exhibits antibacterial activity against ESBL-producing bacteria (5), making it a promising candidate for carbapenem-sparing therapy. Although it is no longer available in many countries, CMZ is available in Japan and is commonly used to treat infections due to ESBLEC, including invasive urinary tract infections (iUTIs). Data on the effectiveness of CMZ for this indication are only available from retrospective single-center or oligo-center studies (6 – 9). Additionally, despite its wide range of approved doses (1–4 g/day in patients with normal renal function, no dosage recommendation in the package insert for renal impairment), there are limited data on appropriate, evidence-based dosing (8, 10). CMZ contains an N-methyl-tetrazole-thiol side chain, which inhibits vitamin K epoxide reductase and may inhibit synthesis of vitamin K-dependent coagulation factors (11). Real-world data on adverse events of CMZ, including coagulopathy, are also scarce (11, 12).
This study aimed to determine the clinical effectiveness of CMZ against iUTI due to ESBLEC compared to MEM. Microbiological characteristics of ESBLEC, optimal dosing of CMZ, and adverse events were also evaluated.
MATERIALS AND METHODS
Study design and patients
This prospective, observational study was conducted at 10 hospitals in Japan between March 2020 and November 2021. Adult patients (age ≥20 y) were eligible for enrollment if they met all three of the following criteria (a, b, c).
(a) Clinical diagnosis of iUTI (a-1 and a-2)
-
(a-1) Clinical symptoms of any of the following
(a-2) Presence of pyuria
-
(b) Microbiological diagnosis
ESBLEC detected in urine culture (≥104 CFU/mL)
If ESBLEC was detected in blood culture only, no other source of infection other than the urinary tract
-
(c) Antibiotic treatment
Either CMZ or MEM initiated within 96 h of culture collection as definitive therapy and continued for at least three calendar days.
Detailed exclusion criteria are available in the supplementary document. The doses of CMZ and MEM were determined at the discretion of each facility.
Clinical data collection and definition
The clinical data were collected from electronic medical records and predefined definitions were used for each condition (supplementary document).
Outcomes
Outcomes, which are detailed in the supplementary material, included clinical and microbiological effectiveness, recurrence within 28 d from the start of either antibiotic, and all-cause mortality (14 d, 30 d, in-hospital). The primary outcome was clinical effectiveness between days 4 and 6 of treatment (early treatment period). Clinical effectiveness was defined as resolution or improvement of clinical symptoms (e.g., fever, low back pain, lateral abdominal pain, renal pain including costovertebral angle tenderness, tachypnea, low blood pressure, and altered mental status) to pre-infection baseline, as determined by an infectious disease specialist (15).
Microbiology
Bacterial identification and susceptibility testing were conducted using MicroScan WalkAway (Beckmann-Coulter, Germany) (six facilities), VITEK2 (bioMérieux, France) (two facilities), and BD phoenix (Beckton Dickinson, USA) (one facility). MALDI Biotyper (Bruker, Bremen, Germany) was also used for bacterial identification in one facility. Bacteria isolated from enrolled patients were sent to the central laboratory based at Kyoto University Graduate School of Medicine for further analysis. At the central laboratory, antibiotic susceptibility was evaluated by broth microdilution (BMD) using a Dry Plate Eiken (Eiken, Tokyo, Japan) according to CLSI guidelines (16). The results were interpreted using the 2021 CLSI breakpoints (16). Detailed microbiological analyses are described in the supplementary material.
Pharmacokinetic/pharmacodynamic analysis
The pharmacokinetic and pharmacodynamic parameters of CMZ and MEM were calculated using the model reported by Tomizawa et al. (17). Creatinine clearance (CrCl) was calculated based on the Cockcroft-Gault equation (18). The time above MIC (TAM) was calculated using the MIC value obtained from each patient and its simulated CMZ and MEM concentration. MIC values determined by BMD at the central laboratory were used for the analysis. Phoenix NLME version 8.1 (Certara, Princeton, NJ, USA) was used for the pharmacokinetic simulation to calculate the TAM.
Statistical analyses
The inverse probability of treatment weight (IPTW) method was used to adjust for baseline confounders of CMZ and MEM treatment (Table S1). Detailed information on statistical analyses is available in the supplementary document.
RESULTS
Comparison of clinical characteristics
Eighty-one and forty-six patients were included in the CMZ and MEM groups, respectively. In univariable analysis, the CMZ group was older, had dependent functional status more often than the MEM group, and had more frequently resided in nursing home or long-term care facilities (LTCF) prior to admission (Table 1). Diabetes was more common in the MEM group than the CMZ group. Use of indwelling device [central venous catheter/central venous port/hemodialysis (CV/HD) catheter, and device other than CV/HD catheter and urinary device] was more prevalent in the MEM group than the CMZ group. The number of days from onset to CMZ or MEM initiation was lower in the MEM group than CMZ group. The MEM group was more likely to be in an intensive care unit (ICU) than the CMZ group on the day of MEM or CMZ initiation. The MEM group had higher qSOFA score, Pitt score, white blood cell count, and C-reactive protein (CRP) level than the CMZ group on the day of MEM or CMZ initiation. The number of days from CMZ or MEM initiation to fever resolution was similar between the two groups [CMZ vs MEM group: 2 d (IQR: 1–6) vs 3 d (1–9)], whereas leukocytosis [CMZ vs MEM group: 3 (4.5%) vs 11 (25.6%)] and high CRP (≥10 mg/dL) [5 (7.6%) vs 16 (37.2%)] were more commonly observed in the early treatment period in the MEM group than the CMZ group (Table S2).
TABLE 1.
Comparison of clinical characteristics of invasive urinary tract infection between cefmetazole and meropenem treatment groups a (n = 127)
| CMZ (n = 81) | MEM (n = 46) | |
|---|---|---|
| Patient demographic and comorbid conditions | ||
| Age | 85 (76–90) | 78 (69–85) |
| Male sex | 28 (34.6) | 23 (50) |
| Dependent functional status | 58 (71.6) | 21 (45.7) |
| Cardiovascular disease | 11 (13.6) | 11 (23.9) |
| Cerebrovascular accident | 25 (31.3) | 10 (21.7) |
| Dementia | 27 (33.3) | 8 (17.4) |
| Connective tissue disease | 6 (7.4) | 9 (19.6) |
| Mild liver disease | 8 (9.9) | 1 (2.2) |
| Moderate to severe kidney disease | 2 (2.5) | 2 (4.3) |
| Diabetes mellitus | 15 (18.5) | 17 (37) |
| Solid tumor (localized) | 13 (16) | 11 (23.9) |
| Solid tumor (metastatic) | 6 (7.4) | 3 (6.5) |
| Leukemia or lymphoma | 0 (0) | 2 (4.3) |
| Charlson comorbidity index | 2 (1–4) | 3 (2-3) |
| Any immunosuppressive status | 6 (7.5) | 8 (17.8) |
| Urological complication | 33 (41.3) | 24 (52.2) |
| Healthcare exposure prior to ESBLEC isolation | ||
| Nursing home or LTCF residence | 32 (40) | 10 (21.7) |
| Hospitalization in the previous 3 mo | 17 (21.3) | 16 (34.8) |
| Hospital onset b | 19 (23.5) | 13 (28.3) |
| Length of hospital stay before isolation of ESBLEC, days | 0 (0–3) | 0 (0–11) |
| Surgery after admission prior to ESBLEC isolation | 5 (6.3) | 6 (13) |
| CV/HD catheter | 0 (0) | 3 (6.5) |
| Urinary device c | 20 (24.7) | 19 (41.3) |
| Device other than CV/HD catheter and urinary device d | 5 (6.2) | 10 (21.7) |
| Any antimicrobial exposure in the previous 1 mo | 27 (33.3) | 15 (32.6) |
| Clinical characteristics and severity of infection | ||
| Bacteremia due to ESBLEC | 35 (43.2) | 27 (58.7) |
| Polymicrobial culture e | 25 (30.9) | 17 (37) |
| Fever f on the day of onset | 72 (90) | 41 (89.1) |
| Pain g on the day of onset | 10 (12.5) | 12 (26.1) |
| Days from onset to microbiological test | 0 (0–1) | 0 (0–2) |
| Days from onset to CMZ or MEM initiation | 2 (0–3) | 0 (0–2) |
| Fever f on the day of CMZ or MEM initiation | 41 (50.6) | 26 (57.8) |
| ICU stay on the day of CMZ or MEM initiation | 1 (1.2) | 8 (17.4) |
| qSOFA on the day of CMZ or MEM initiation | 0 (0–1) | 1 (0–2) |
| Pitt bacteremia score h on the day of CMZ or MEM initiation | 3 (0–3) | 3 (3–5) |
| WBC on the day of CMZ or MEM initiation (/µL) | 8,900 (6,563–12,598) | 10,825 (7,405–16,318) |
| Leukocytosis i on the day of CMZ or MEM initiation | 19 (27.1) | 17 (40.5) |
| CRP on the day of CMZ or MEM initiation (mg/dL) | 7.3 (4.2–14.5) | 13.2 (5.5–23.8) |
| CRP >10 (mg/dL) | 30 (42.9) | 27 (64.3) |
Data are presented as number (%) or median (interquartile range) unless indicated otherwise. Abbreviations. CMZ, cefmetazole; CV, central venous catheter/central venous port; ESBLEC, ESBL–producing E. coli; HD, hemodialysis; ICU, intensive care unit; LTCF, long-term care facilities, MEM, meropenem, WBC, white blood cell.
Defined as length of hospital stay before isolation of ESBLEC equal or longer than 4 d.
Urinary device includes urinary catheter, ureteral stent, and nephrostomy catheter.
Includes percutaneous endoscopic gastrostomy tube, tracheostomy tube, endotracheal tube, and nasogastric tube.
Isolation of additional bacteria other than ESBLEC from the same culture.
Fever is defined as temperature equal or higher than 37.5ºC.
Pain includes lower back pain, lateral abdominal pain, and renal pain (including costovertebral angle tenderness).
Pitt bacteremia score is calculated only for bacteremic cases.
Leukocytosis is defined as WBC>12000 μL.
Antibiotic treatment
The antibiotic treatment given is summarized in Table 2. Total duration of study drug as well as total duration of all antibiotics were similar between the two groups. Empirical antibiotic treatment was used more frequently in the CMZ group than the MEM group, with ceftriaxone use significantly more common in the CMZ group than the MEM group. No statistical difference was noted in the empiric use of potentially effective antibiotics against ESBLEC, such as piperacillin-tazobactam (TZP). MEM was used in three cases (3.7%) of the CMZ group as empiric therapy.
TABLE 2.
Details of antibiotic treatment a
| Variable | CMZ (n = 81) | MEM (n = 46) | P value b |
|---|---|---|---|
| Total duration of study drug, median (IQR) | 8 (6–12) | 9 (5–12) | 0.920 |
| Total duration of all antibiotics, median (IQR) | 11 (8–14) | 12 (10–16) | 0.087 |
| Empiric therapy c | |||
| None | 24 (29.6%) | 28 (60.9%) | 0.001 |
| AMP | 1 (1.2%) | 0 | >0.999 |
| SAM | 8 (9.9%) | 3 (6.5%) | 0.745 |
| TZP | 9 (11.1%) | 4 (8.7%) | 0.768 |
| CFZ | 1 (1.2%) | 0 | >0.999 |
| CRO | 33 (40.7%) | 8 (17.4%) | 0.01 |
| FEP | 1 (1.2%) | 1 (2.2%) | >0.999 |
| MEM | 3 (3.7%) | NA | NA |
| FQ d | 2 (2.5%) | 2 (4.3%) | 0.62 |
| Oral cephalosporin e | 2 (2.5%) | 0 | 0.534 |
| Treatment continuation after study therapy f | |||
| Any | 16 (19.8%) | 24 (52.2%) | <0.001 |
| SAM | 1 (1.2%) | 0 | >0.999 |
| AMC | 0 | 1 (2.2%) | 0.362 |
| TZP | 1 (1.2%) | 0 | >0.999 |
| CMZ | NA | 16 (34.8%) | NA |
| MEM | 2 (2.5%) | NA | NA |
| FQ d | 4 (4.9%) | 1 (2.2%) | 0.653 |
| SXT | 7 (8.6%) | 1 (2.2%) | 0.257 |
| Other g | 1 (1.2%) | 2 (4.3%) | 0.297 |
Abbreviations. AMC, amoxicillin–clavulanic acid; AMP, ampicillin; CFZ, cefazolin; CRO, ceftriaxone; FEP, cefepime; FQ, fluoroquinolone (levofloxacin, ciprofloxacin, garenoxacin); NA, not available; SAM, ampicillin–sulbactam; SXT, trimethoprim–sulfamethoxazole; TZP, piperacillin–tazobactam.
Bold font indicates statistically significant results (P <0.05).
Empiric therapy was defined as antibiotics active against Gram-negative bacteria that have been used within 96 h prior to the start of the study drug (i.e., CMZ or MEM). Four patients (three in CMZ group and one in MEM group) received two antibiotics, and both antibiotics were counted separately.
FQ include two levofloxacin, one ciprofloxacin, and one garenoxacin in empiric treatment, and three levofloxacin and one ciprofloxacin in treatment continuation.
Oral cephalosporin includes one cefditoren pivoxil and one cefcapene pivoxil.
Two patients (one in CMZ group and one in MEM group) received two antibiotics, and both antibiotics were counted separately for each antibiotics category.
Includes one patient each treated with fosfomycin and minocycline (MEM group) and one patient who received CFZ (CMZ group).
Switch from definitive therapy with CMZ or MEM to another antibiotic therapy was more frequent in the MEM group than the CMZ group (P < 0.001). No significant differences were found between the two groups regarding the individual antimicrobial agents switched. CMZ replaced MEM in 34.5% (n = 16) of the MEM group. For two patients who received MEM after CMZ, one patient clinically improved but was changed to MEM due to liver dysfunction, and in the other, CMZ was clinically and microbiologically effective prior to the switch to MEM.
MIC and susceptibility of isolated ESBLEC
All tested isolates (n = 124) were susceptible to MEM with low MIC (≤0.12 mg/L). The MICs to CMZ ranged from ≤1 (n = 86) to 8 mg/L (n = 5) (Table 3). Susceptibility rates to β-lactam/β-lactamase inhibitors, such as amoxicillin-clavulanate and TZP, were 77.4% (n = 96) and 96% (n = 119), respectively. The highest MIC of TZP was 16 mg/L [n = 5 4%)]. Susceptibility to ciprofloxacin was as low as 15.3% (n = 19), and approximately half (n = 66, 53.2%) of the isolates were resistant to trimethoprim-sulfamethoxazole. Among aminoglycosides, amikacin was the most active agent with all 124 isolates testing susceptible.
TABLE 3.
MIC and susceptibility of isolated ESBLEC a (mg/L) (n = 124)
| AMC | TZP c | CFZ d | CTX | CAZ | FEP | CMZ e | FMOX | ATM | |
|---|---|---|---|---|---|---|---|---|---|
| MIC50 | 8/4 | 2/4 | >8 | >16 | 4 | 8 | <1 | <0.12 | 8 |
| MIC90 | 16/8 | 8/4 | >8 | >16 | 16 | >32 | 4 | 0.25 | >16 |
| Susceptible, number b , (%) | 96 (77.4%) | 119 (96%) | 0 | 1 (0.8%) |
63 (50.8%) | 29 (23.4%) | 124 (100%) | 31 (25%) |
|
| MEM | IMP | FRPM | GEN | TOB | AMK | CIP | SXT | FOF f | |
| MIC50 | ≤0.12 | <1 | 1 | <4 | <4 | ≤16 | >４ | <2/38 | ≤16 |
| MIC90 | ≤0.12 | <1 | 2 | >16 | 16 | ≤16 | >4 | >4/76 | ≤16 |
| Susceptible, number, (%) | 124 (100%) | 124 (100%) | 99 (79.8%) | 99 (79.8%) | 124 (100%) | 19 (15.3%) | 66 (53.2%) | 120 (96.8%) |
Abbreviations. AMK, amikacin; AMC, amoxicillin–clavulanic acid; ATM, aztreonam; CAZ, ceftazidime; CFZ, cefazolin; CIP, ciprofloxacin; CTX, cefotaxime; FEP, cefepime; FMOX, flomoxef; FOF, fosfomycin; FRPM, faropenem; GEN, gentamicin; SXT, trimethoprim-sulfamethoxazole; TOB, tobramycin; TZP. piperacillin-tazobactam.
Susceptibility results were interpreted using the 2021 CLSI breakpoints.
TZP breakpoint MIC < 8/4 mg/L was used for susceptible category.
CFZ breakpoint MIC < 2 mg/L was used for susceptible category.
MIC distribution of CMZ is as follows: <1 mg/L (n=86, 69.4%), 2 mg/L (n=23, 18.5%), 4 mg/L (n=10, 8%), 8 mg/L (n=5, 4%).
Fosfomycin MICs were measured by broth microdilution in the presence of glucose-6-phosphate (25 mg/L in the medium).
Molecular characteristics of isolated ESBLEC
In all isolates, bla CTX-M was detected as the ESBL gene (Table 4). The predominant CTX-M subtype was CTX-M-27 (n = 59, 47.6%), followed by CTX-M-15 (n = 30, 24.2%) and CTX-M-14 (n = 25, 20.2%). ST131 accounted for 73.4% (n = 91) of the clones, followed by ST1193 (n = 6, 4.8%), and the rest comprised 18 different sequence types (STs). ST131 clades included C1-M27 (n = 43, 47.3%), C1-non-M27 (n = 20, 22%), C2 (n = 17, 18.7%), and A (n = 8, 8.8%).
TABLE 4.
Molecular characteristics of isolated ESBL-producing E. coli (n = 124)
| MLST_ST a | CTX-M subtype | ST131 clade (n = 91) | |||
|---|---|---|---|---|---|
| 131 | 91 (73.4%) | 27 | 59 (47.6%) | C1-M27 | 43 (47.3%) |
| 1193 | 6 (4.8%) | 15 | 30 (24.2%) | C-non-M27 | 20 (22%) |
| 38 b | 6 (4.8%) | 14 | 25 (20.2%) | C2 | 17 (18.7%) |
| 95 | 3 (2.4%) | 55 | 4 (3.2%) | A | 8 (8.8%) |
| 10 | 2 (1.6%) | 8 | 2 (1.6%) | B | 2 (2.2%) |
| 69 | 2 (1.6%) | 65 | 2 (1.6%) | C0 | 1 (1.1%) |
| 393 | 2 (1.6%) | 3 | 1 (0.8%) | ||
| 104 | 1 (0.8%) | ||||
n=1 (0.8%) for ST12, 23, 73, 155, 162, 215, 450, 533, 648, 803, 1588, and 5150, respectively.
includes one isolate with single locus variant of ST38.
Genes for the following other β-lactamases were detected: TEM-1, 26 (21%); OXA-1, 14 (11.3%); TEM-190, 1 (0.8%); TEM-135, 1 (0.8%). Genes for pAmpC were detected in only one isolate (DHA-1), and the CMZ MIC of the isolate was ≤1 mg/L.
Outcomes
The all-cause 30 d and in-hospital mortality rates were higher in the MEM group than the CMZ group [n = 5 (12.5%) vs 0, n = 6 (13.3%) vs n = 2 (2.6%), respectively (Table 5)]. The median days from the completion of the treatment to the recurrence was 11 d (range, 6–27) in the CMZ group and 11 d (range, 8–14) in the MEM group. In all cases with follow-up urine cultures, both drugs were microbiologically effective. Clostridioides difficile infection (CDI) within 4 wk after the end of treatment was observed in two (2.5%) patients in the CMZ group and one (2.2%) patient in the MEM group. Days from CMZ treatment to CDI onset were 4 d and 26 d in the two CMZ patients, respectively, and the onset of CDI could not be determined in the patient in the MEM group. Carbapenem-resistant Enterobacterales was not detected in any clinical specimens within 4 wk of the end of treatment in either group.
TABLE 5.
Comparison of outcomes of invasive urinary tract infection between cefmetazole and meropenem treatment groups a (n = 127)
| CMZ (n = 81) | MEM (n = 46) | |
|---|---|---|
| Outcome | ||
| Clinically effective (early) | 75 (92.6) | 38 (82.6) |
| Clinically effective (late) b | 73 (96.1) | 40 (90.9) |
| Microbiologically effective (early) c | 61 (100) | 22 (100) |
| Microbiologically effective (late) c | 9 (100) | 5 (100) |
| 14-d mortality d | 0 (0) | 1 (2.3) |
| 30-d mortality d | 0 (0) | 5 (12.5) |
| In-hospital mortality d | 2 (2.6) | 6 (13.3) |
| Recurrence within 28 d | 6 (8.3) | 2 (5.6) |
| LOS after isolation of ESBLEC among survivors, days | 15 (11–34) | 19 (14–35) |
| Days from CMZ or MEM initiation to fever resolution e | 2 (1–6) | 3 (1–9) |
| Defervescence from CMZ or MEM initiation to early treatment period | 29 (39.7) | 14 (31.8) |
| Inadequate source control | 4 (4.9) | 4 (8.7) |
| Dependent functional status on discharge | 52 (68.4) | 23 (57.5) |
| Discharge to home | 36 (46.2) | 22 (55) |
| Clostridiodes difficile infection within 28 d after treatment | 2 (2.5) | 1 (2.2) |
| CRE isolation within 28 d after treatment | 0 (0) | 0 (0) |
Data are presented as number (%) or median (interquartile range) unless indicated otherwise. "Early" indicates early treatment period, and "late" indicates late treatment period. Abbreviations. CMZ, cefmetazole; CRE, carbapenem-resistant Enterobacterales; ESBLEC, ESBL–producing E. coli; LOS, length of hospital stay; MEM, meropenem.
Data are missing for five (CMZ) and two (MEM) cases.
Data are available for 83 (CMZ 61. MEM 22) (early) and 14 (CMZ 9 MEM 5) (late) cases.
Data are missing for 6/2 (CMZ/MEM) cases for 14-d mortality, 16/6 (CMZ/MEM) cases for 30-d mortality, and 4/1 (CMZ/MEM) cases for in-hospital mortality.
Fever is defined as temperature equal or higher than 37.5ºC.
Mortality rates for bacteremic patients due to ESBLEC were similar between the two groups: 14 d mortality, 0 vs 0; 30 d mortality, 0 vs 2 (9.1%); in-hospital mortality, 0 vs 2 (7.4%) for the CMZ and MEM groups, respectively.
After PS adjustment, clinical effectiveness did not differ between the two groups in both the early [adjusted odds ratio (aOR): 0.63 (95% CI: 0.15–2.75), P = 0.54] and late treatment periods [aOR: 1.83 (0.26–12.75), P = 0.54] (Table 6). Adjusted odds ratio for 14 d mortality was not available due to the small number of events. The risk of 30 d mortality was lower in CMZ group [aOR: <0.01 (95% CI: NA), P < 0.01], whereas the risk of in-hospital mortality, recurrence, and LOS after isolation of ESBLEC among survivor were similar in both groups [aOR: 0.20 (0.03–1.36), P = 0.10; aOR: 1.36 (0.14–12.93), P = 0.79; P = 0.23, respectively].
TABLE 6.
Propensity score-adjusted analyses of clinical outcomes of invasive UTI: cefmetazole vs meropenem treatment groups a
| Variables | Adjusted odds ratio (95% CI) | P-value |
|---|---|---|
| Clinically effective (early) | 0.63 (0.15–2.75) | 0.54 |
| Clinically effective (late) | 1.83 (0.26–12.75) | 0.54 |
| 14 d mortality | NA | NA |
| 30 d mortality | <0.01 (NA) | <0.01 |
| In-hospital mortality | 0.20 (0.03–1.36) | 0.10 |
| Recurrence within 28 d | 1.36 (0.14–12.93) | 0.79 |
| LOS after isolation of ESBLEC among survivors | NA | 0.23 |
The propensity score was calculated using a nonparsimonious multivariable logistic regression model including the baseline characteristic variables (age, sex, nursing home or LTCF residence, hospitalization in the past 3 mo, surgery after admission prior to ESBLEC isolation, hospital onset, ESBLEC bacteremia, polymicrobial isolation, Charlson comorbidity index, immunocompromised status, urological complication, device other than CV/HD catheter and urinary device, qSOFA score, any empiric therapy before initiation of CMZ or MEM, and CRP >10 mg/dL [(as defined in Table 1). The adjusted odds ratio for 14-d mortality is not available owing to the small number of events. Abbreviations. LOS, length of stay; NA, not available.
Pharmacokinetic/pharmacodynamic analysis
Dosing of CMZ and MEM with TAM is summarized in Table 7. Except for one patient in the CMZ group with CrCl of 32 mL/min, for whom CMZ (2 g q 24 h) was infused over 120 min, CMZ and MEM were infused over 30 min [CMZ: 19 (23.5%), MEM: 7 (15.2%)] or 60 min [CMZ: 61 (75.3%), MEM: 39 (84.8%)]. In the MEM group, TAM was 100% in all cases. In all six cases where CMZ was clinically ineffective in the early treatment period, TAM of CMZ was 100%.
TABLE 7.
Summary of cefmetazole or meropenem dosing in patients with invasive urinary tract infection due to ESBL-producing E. coli a
| CMZ (n = 81) | MEM (n = 46) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CrCl (mL/min) category | n (%) | Dose (g) | Interval (h) | No. | MIC, median (range) | TAM for each regimen, median (IQR) | Clinically ineffective (early), no. | CrCl (mL/min) category | n (%) | Dose (g) | Interval (hour) | No. | Clinically ineffective (early), no. |
| <10 | 3 (3.7%) | 0.5 | 12 | 1 | <1 | 100 | 1 | <10 | 5 (11%) | 0.5 | 24 | 4 | |
| 1 | 24 | 1 | 4 | 99.8 | 1 | 24 | 1 | ||||||
| 2 | 24 | 1 | <1 | 100 | 10–25 | 16 (35%) | 1 | 12 | 6 | 1 | |||
| 10–29 | 20 (24.7%) | 1 | 12 | 11 | <1 (<1–4) | 100 (100–100) |
0.5 | 12 | 5 | ||||
| 1 | 24 | 5 b | NA (<1–4) | 80.9 (64.7–96) |
1 | 8 | 2 | ||||||
| 2 | 12 | 3 | 2 (<1–8) | 100 (100–100) |
1 | 0.5 | 8 | 1 | |||||
| 2 | 24 | 1 | 2 | 82.9 | 0.5 | 24 | 1 | ||||||
| 30–50 | 29 (35.8%) | 1 | 8 | 8 | <1 (<1–2) | 100 (100–100) |
4 | 0.25 | 12 | 1 | |||
| 1 | 12 | 9 | <1 (<1–4) | 100 (84.6–100) |
26–50 | 11 (24%) | 1 | 12 | 6 | 2 | |||
| 1 | 24 | 5 | <1 (<1–2) | 65 (55.6–90.2) |
1 | 8 | 2 | 1 | |||||
| 2 | 12 | 1 | 2 | 100 | 0.5 | 12 | 2 | ||||||
| 2 | 24 | 6 | <1 (<1–8) | 58.5 (43.0–62.6) |
2 | 12 | 1 | ||||||
| >50 | 29 (35.8%) | 1 | 6 | 2 | <1 (<1–<1) | 100 (100–100) |
>50 | 14 (30%) | 1 | 8 | 9 | 2 | |
| 1 | 8 | 10 b | <1 (<1–2) | 50 | 0.5 | 12 | 2 b | 2 | |||||
| 1 | 12 | 11 | <1 (<1–8) | 57.8 (41.5–82.4) |
0.5 | 6 | 1 | ||||||
| 2 | 12 | 4 | <1 (<1–<1) | 95.2 (89.6–100) |
0.5 | 8 | 1 | ||||||
| 2 | 24 | 2 | NA (2–8) | 43.6 (32.8–NA) | 1 | 12 | 1 | ||||||
Abbreviations. NA, not available; TAM, time above MIC.
For one case per each category, TAM could not be calculated; two isolates were missing from microbiological analyses at the central laboratory. One isolate was not identified as ESBL-producing E. coli in the central laboratory analysis, and thus excluded from the analysis (ESBL production of the E. coli isolate was reported in the hospital microbiological laboratory, with resistance to cefotaxime).
Adverse events
Regarding adverse events considered related to CMZ or MEM, three patients in the CMZ group had liver dysfunction (Grade 1: n = 2, Grade 2: n = 1) and one patient had skin rash (Grade 1); one patient in the MEM group had non-CDI diarrhea (Grade 4).
Data were available for 35 patients (CMZ: 18, MEM: 17) for prothrombin time (PT) and 33 patients (CMZ: 17, MEM: 16) for activated partial thromboplastin time (APTT) in the early treatment period. Compared to the pre-treatment period, no significant increase was observed for PT [median increase presented as fold change (IQR); CMZ: 1 (0.8–1.1), MEM 0.9 (0.8–1), P = 0.291] or APTT [CMZ: 0.9 (0.8–1.1), MEM 0.9 (0.8–1.2), P = 0.801] in either group in the early treatment period. Data were available for 19 patients (CMZ: 14, MEM: 5) for PT and 19 APTT (CMZ: 14, MEM: 5) in the late treatment period. Compared to pre-treatment, there was no significant increase for PT [CMZ: 1 (0.9–1), MEM 0.9 (0.7–1.1), P-value 0.781] or APTT [CMZ: 1 (0.9–1.1), MEM 1.2 (1–1.6), P = 0.064]) in either group in the late treatment period.
DISCUSSION
This observational study showed that CMZ was as effective as MEM in a cohort of patients with iUTIs caused by ESBLEC. This study includes the largest number of patients and was conducted at the largest number of facilities among studies that examined the effectiveness of CMZ for ESBLEC to date (6 – 9). Among previous reports, three studies compared the effectiveness of CMZ with that of carbapenems, and in all three studies CMZ was as effective as carbapenems. However, only one study adjusted for background factors using propensity scores (9). In that study focusing on ESBLEC bloodstream infection (BSI) (hereafter referred to as CF-CARBA study) (9), patients treated with CMZ were included as part of the combined CMZ and flomoxef group (CF), where flomoxef is an oxacephem agent with activity against ESBLEC that is approved for clinical use in Japan. The unadjusted 30 d mortality rates in the definitive treatment cohort were 9.3% in the carbapenem group and 5.1% in the CF group, and 7.0% and 7.4%, respectively, after PS adjustment. The median age of patients was approximately 10 y higher in the present study than the CF-CARBA study. The CF-CARBA study included more immunocompromised patients. The present study also included patients other than those with bacteremia, and the 30 d unadjusted mortality rate was lower in the CMZ group than in the CF-CARBA study, but higher in the MEM group than the CF-CARBA study. Intriguingly, the 30 d mortality rate in the MEM group was lower when only the patients with BSI due to ESBLEC were included, to almost the same level as in the CF-CARBA study. Although identifying the cause of death was outside this study’s scope, it is possible that causes of death not directly related to infection were also involved, which may have contributed to the lower mortality associated with the CMZ group than the MEM group, even after PS adjustment. Despite the use of slightly different criteria, the two studies were consistent in demonstrating high clinical effectiveness of CMZ against ESBLEC infection.
The dosing of CMZ was also examined in this study. In this cohort, the MIC of CMZ for all ESBLEC strains available for analysis (n = 124) was ≤8 mg/L. In the majority of cases, a dose of 1 g q 8 h for patients with CrCl >50 mL/min and 1 g q 12 h for CrCl 30–50 mL/min resulted in a TAM >50%, which is the pharmacodynamic target associated with effectiveness in the treatment of ESBLEC infections (19). Clinical failure occurred in patients with high TAM in both the MEM and CMZ groups, which is consistent with our previous report (8) and may be more related to the patients’ underlying conditions than the intrinsic efficacy of the antibiotics. Of the six patients in the CMZ group and the eight patients in the MEM group for whom study drugs were clinically ineffective, microbiologically effectiveness was confirmed in all patients for whom microbiological evaluation was possible (four patients in the CMZ group and six patients in the MEM group). There was no clear difference between CMZ and MEM in safety, including coagulation function tests. CMZ is reportedly associated with coagulopathy and increased hemorrhagic events (11, 12), but we did not observe a signal in our relatively small cohort.
We also performed a molecular microbiological analysis. In concordance with the global distribution, ST131 was the major ST in this study. However, nearly half of the CTX-M gene subtypes were CTX-M-27. CTX-M-14, which differs from CTX-M-27 by only one nucleotide, accounted for 20%. The global pandemic subtype CTX-M-15 accounted for only 24%, and the ST131 clade C1-M27, accounted for approximately 50%, confirming a unique genetic background of ESBLECs in Japan (20). These molecular epidemiological differences may explain the much lower detection rate of bla OXA-1 (11.3%) in this cohort compared to the MERINO trial cohort (67.6%) (4). As the presence of bla OXA-1 is associated with higher MICs of TZP (21), our study suggests that the results of the MERINO trial may not be directly applicable in Japan.
CMZ is hydrolyzed by AmpC and strains producing this group of enzymes are resistant to this agent. In the present study, pAmpC was detected in only 0.8% (n = 1). In a previous report from Japan, co-production of ESBL and pAmpC genes was relatively rare (5). 3GCR E. coli with a CMZ MIC of 16 mg/L are considered susceptible by CLSI criteria (16), but increasing rates of pAmpC production among E. coli strains with CMZ MICs of 16 mg/L or higher have been reported (5), which should be noted when interpreting MIC data.
The antimicrobial agents prescribed prior to study drug administration differed between the two groups, with MEM initiated more often without prior drug administration and CMZ administered more frequently after other drugs, especially ceftriaxone. The propensity to receive TZP, which might still be effective for ESBLEC, did not differ between the two groups. The patients in the MEM group tended to be more ill at the initiation of the study drug administration than the CMZ group (e.g., higher qSOFA score, Pitt bacteremia score, and CRP levels; more patients in ICU). These differences were considered when adjusting for PS. Approximately one-third of patients in the MEM group were later de-escalated to CMZ to complete the treatment. Since the switch occurred when treatment of the acute phase of infection had already been completed, it is unlikely to have affected the study outcome.
This study has certain limitations. Due to the observational nature of the study, there were some missing data regarding some parameters including microbiological effectiveness. The significant difference in 30 d mortality does not reflect superiority of CMZ over MEM, but rather the fact that there were few deaths in the CMZ group. Although PS was used to minimize the baseline differences between the two groups, there may be residual differences that could not be fully adjusted as the mortality risk at baseline seemed to differ substantially between the two groups.
This study has shown that the clinical and bacteriological effectiveness of CMZ was comparable to MEM in the treatment of iUTI cases caused by ESBLEC in a cohort that included more than 40% of patients with concomitant bacteremia. In addition, an appropriate dosing strategy of CMZ was developed. Based on these findings, CMZ appears to be a safe and effective alternative to MEM without increase in mortality, however, a randomized clinical trial (RCT) is necessary to conclusively demonstrate this; one is currently in progress (22).
ACKNOWLEDGMENTS
This work was supported by a Grant-in-Aid for Scientific Research (C) (grant no. 19K10573) (K. H.); Grant-in-Aid for Scientific Research (B) (grant no. 22H03323) (Y. D.); and a grant for International Health Research from the Ministry of Health, Labor and Welfare of Japan (grant no. 19A1022) (K. H.).
KH, YM, AS, RT, KS, TH, RH, TM, HK, MM, HH, SS, and YD contributed to the conception of the idea of the study and collection of data. KU and ST conducted statistical analyses. YH conducted PK/PD analyses. YM conducted microbiological analyses. KH drafted the original manuscript. NO supervised the conduct of this study. All authors reviewed the manuscript draft and revised it critically on intellectual content. All authors approved the final version of the manuscript to be published.
The authors declare that there is no conflict of interest.
Contributor Information
Kayoko Hayakawa, Email: khayakawa@hosp.ncgm.go.jp.
Laurent Poirel, University of Fribourg, Fribourg, Switzerland .
ETHICS APPROVAL
The study was approved by the institutional review board at the National Center for Global Health and Medicine (No. NCGM-G-003389-00). The opt-out recruitment method was used, and the requirement for informed consent was waived.
DATA AVAILABILITY
Data are available upon reasonable request with the permission of participating facilities. The sequences determined in this study have been deposited in the GenBank/ENA/DDBJ Sequence Read Archive database (PRJNA976817).
SUPPLEMENTAL MATERIAL
The following material is available online at https://doi.org/10.1128/aac.00510-23.
Supplemental methods & Tables S1 and S2.
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
REFERENCES
- 1. WHO . 2021. GLASS report. https://www.who.int/publications/i/item/9789240027336.
- 2. JANIS . 2020. Annual open report 2020. Available from: https://janis.mhlw.go.jp/english/report/open_report/2020/3/1/ken_Open_Report_Eng_202000_clsi2012.pdf
- 3. Ahn JY, Song JE, Kim MH, Choi H, Kim JK, Ann HW, Kim JH, Jeon Y, Jeong SJ, Kim SB, Ku NS, Han SH, Song YG, Yong D, Lee K, Kim JM, Choi JY. 2014. Risk factors for the acquisition of carbapenem-resistant Escherichia coli at a tertiary care center in South Korea: a matched case-control study. Am J Infect Control 42:621–625. doi: 10.1016/j.ajic.2014.02.024 [DOI] [PubMed] [Google Scholar]
- 4. Harris PNA, Tambyah PA, Lye DC, Mo Y, Lee TH, Yilmaz M, Alenazi TH, Arabi Y, Falcone M, Bassetti M, Righi E, Rogers BA, Kanj S, Bhally H, Iredell J, Mendelson M, Boyles TH, Looke D, Miyakis S, Walls G, Al Khamis M, Zikri A, Crowe A, Ingram P, Daneman N, Griffin P, Athan E, Lorenc P, Baker P, Roberts L, Beatson SA, Peleg AY, Harris-Brown T, Paterson DL, MERINO Trial Investigators and the Australasian Society for Infectious Disease Clinical Research Network (ASID-CRN) . 2018. Effect of piperacillin-tazobactam vs meropenem on 30-day mortality for patients with E coli or Klebsiella pneumoniae bloodstream infection and ceftriaxone resistance: a randomized clinical trial. JAMA 320:984–994. doi: 10.1001/jama.2018.12163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Matsumura Y, Yamamoto M, Nagao M, Tanaka M, Takakura S, Ichiyama S. 2016. In vitro activities and detection performances of cefmetazole and flomoxef for extended-spectrum β-lactamase and plasmid-mediated AmpC β-lactamase-producing Enterobacteriaceae. Diagn Microbiol Infect Dis 84:322–327. doi: 10.1016/j.diagmicrobio.2015.12.001 [DOI] [PubMed] [Google Scholar]
- 6. Doi A, Shimada T, Harada S, Iwata K, Kamiya T. 2013. The efficacy of cefmetazole against pyelonephritis caused by extended-spectrum β-lactamase-producing Enterobacteriaceae. Int J Infect Dis 17:e159–63. doi: 10.1016/j.ijid.2012.09.010 [DOI] [PubMed] [Google Scholar]
- 7. Fukuchi T, Iwata K, Kobayashi S, Nakamura T, Ohji G. 2016. Cefmetazole for bacteremia caused by ESBL-producing Enterobacteriaceae comparing with carbapenems. BMC Infect Dis 16:427. doi: 10.1186/s12879-016-1770-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hamada Y, Matsumura Y, Nagashima M, Akazawa T, Doi Y, Hayakawa K. 2021. Retrospective evaluation of appropriate dosing of cefmetazole for invasive urinary tract infection due to extended-spectrum β-lactamase-producing Escherichia coli. J Infect Chemother 27:1602–1606. doi: 10.1016/j.jiac.2021.07.009 [DOI] [PubMed] [Google Scholar]
- 9. Matsumura Y, Yamamoto M, Nagao M, Komori T, Fujita N, Hayashi A, Shimizu T, Watanabe H, Doi S, Tanaka M, Takakura S, Ichiyama S. 2015. Multicenter retrospective study of cefmetazole and flomoxef for treatment of extended-spectrum-β-lactamase-producing Escherichia coli bacteremia. Antimicrob Agents Chemother 59:5107–5113. doi: 10.1128/AAC.00701-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Hamada Y, Kasai H, Suzuki-Ito M, Matsumura Y, Doi Y, Hayakawa K. 2022. Pharmacokinetic/pharmacodynamic analysis and dose optimization of cefmetazole and flomoxef against extended-spectrum β-lactamase-producing enterobacterales in patients with invasive urinary tract infection considering renal function. Antibiotics 11:456. doi: 10.3390/antibiotics11040456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Breen GA, St Peter WL. 1997. Hypoprothrombinemia associated with cefmetazole. Ann Pharmacother 31:180–184. doi: 10.1177/106002809703100210 [DOI] [PubMed] [Google Scholar]
- 12. Chen L-J, Hsiao F-Y, Shen L-J, Wu F-L, Tsay W, Hung C-C, Lin S-W, Arruda VR. 2016. Use of hypoprothrombinemia-inducing cephalosporins and the risk of hemorrhagic events: a nationwide nested case-control study. PLoS ONE 11:e0158407. doi: 10.1371/journal.pone.0158407 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Velasco M, Horcajada JP, Mensa J, Moreno-Martinez A, Vila J, Martinez JA, Ruiz J, Barranco M, Roig G, Soriano E. 2001. Decreased invasive capacity of quinolone-resistant Escherichia coli in patients with urinary tract infections. Clin Infect Dis 33:1682–1686. doi: 10.1086/323810 [DOI] [PubMed] [Google Scholar]
- 14. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche J-D, Coopersmith CM, Hotchkiss RS, Levy MM, Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent J-L, Angus DC. 2016. The third international consensus definitions for sepsis and septic shock (sepsis-3). JAMA 315:801–810. doi: 10.1001/jama.2016.0287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Rubin RH, Shapiro ED, Andriole VT, Davis RJ, Stamm WE. 1992. Evaluation of new anti-infective drugs for the treatment of urinary tract infection. Clin Infect Dis 15:S216–S227. doi: 10.1093/clind/15.Supplement_1.S216 [DOI] [PubMed] [Google Scholar]
- 16. CLSI . 2021. M100. Performance standards for antimicrobial susceptibility testing. Clinical & Laboratory Standards Institute. [Google Scholar]
- 17. Tomizawa A, Nakamura T, Komatsu T, Inano H, Kondo R, Watanabe M, Atsuda K. 2017. Optimal dosage of cefmetazole for intraoperative antimicrobial prophylaxis in patients undergoing surgery for colorectal cancer. J Pharm Health Care Sci 3:1. doi: 10.1186/s40780-016-0071-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Cockcroft DW, Gault MH. 1976. Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41. doi: 10.1159/000180580 [DOI] [PubMed] [Google Scholar]
- 19. Andes D, Craig WA. 2005. Treatment of infections with ESBL-producing organisms: pharmacokinetic and pharmacodynamic considerations. Clin Microbiol Infect 11:10–17. doi: 10.1111/j.1469-0691.2005.01265.x [DOI] [PubMed] [Google Scholar]
- 20. Matsumura Y, Pitout JDD, Gomi R, Matsuda T, Noguchi T, Yamamoto M, Peirano G, DeVinney R, Bradford PA, Motyl MR, Tanaka M, Nagao M, Takakura S, Ichiyama S. 2016. Global Escherichia coli sequence type 131 clade with blaCTX-M-27 gene. Emerg Infect Dis 22:1900–1907. doi: 10.3201/eid2211.160519 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Henderson A, Paterson DL, Chatfield MD, Tambyah PA, Lye DC, De PP, Lin RTP, Chew KL, Yin M, Lee TH, Yilmaz M, Cakmak R, Alenazi TH, Arabi YM, Falcone M, Bassetti M, Righi E, Rogers BA, Kanj SS, Bhally H, Iredell J, Mendelson M, Boyles TH, Looke DFM, Runnegar NJ, Miyakis S, Walls G, Khamis MAI, Zikri A, Crowe A, Ingram PR, Daneman N, Griffin P, Athan E, Roberts L, Beatson SA, Peleg AY, Cottrell K, Bauer MJ, Tan E, Chaw K, Nimmo GR, Harris-Brown T, Harris PNA, MERINO Trial Investigators and the Australasian Society for Infectious Disease Clinical Research Network (ASID-CRN) . 2021. Association between minimum inhibitory concentration, β-lactamase genes and mortality for patients treated with piperacillin/tazobactam or meropenem from the MERINO study. Clin Infect Dis 73:e3842–e3850. doi: 10.1093/cid/ciaa1479 [DOI] [PubMed] [Google Scholar]
- 22. Hayakawa K, Yohei D, et al. 2022. Randomised controlled trial of meropenem versus cefmetazole for definitive treatment of bloodstream infections due to ESBL. Available from: https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000055809
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental methods & Tables S1 and S2.
Data Availability Statement
Data are available upon reasonable request with the permission of participating facilities. The sequences determined in this study have been deposited in the GenBank/ENA/DDBJ Sequence Read Archive database (PRJNA976817).