Efficacy and Safety of Cefmetazole for Bacteremia Caused by Extended-Spectrum β-Lactamase–Producing Enterobacterales vs Carbapenems: A Retrospective Study

Abstract Background Extended-spectrum β-lactamase (ESBL)–producing Enterobacterales have become a global concern owing to increased infections, high mortality, and limited antibiotic treatment options. Carbapenems (CPMs) are effective against ESBL-producing Enterobacterales, but their overuse leads to the emergence of multidrug-resistant bacteria. Cefmetazole (CMZ) is effective in vitro; however, its clinical efficacy remains unclear. Methods We retrospectively reviewed patients who were treated with CMZ or CPMs for bacteremia caused by ESBL-producing Enterobacterales between 1 April 2014 and 31 September 2022 at Tenri Hospital. The primary outcome measure was 90-day mortality. We also evaluated resistance genes and sequence types of ESBL-producing Enterobacterales. Results In total, 156 patients were enrolled in this study. Ninety patients (58%) received CMZ therapy. Patients in the CMZ group were significantly older than those in the CPM group (median [IQR], 79 years [71–86] vs 74 years [64–83]; P = .001). The severity of the Pitt bacteremia score of the CMZ group was lower than that in the CPM group (0 [0–2] vs 2 [0–2], P = .042). Six patients (7%) in the CMZ group and 10 (15%) in the CPM group died by day 90 (P = .110). Charlson Comorbidity Index and prevalence of sequence 131 between the groups were statistically insignificant. Conclusions Our findings suggest that CMZ is a well-tolerated alternative to CPM for treating bacteremia caused by ESBL-producing Enterobacterales.

The number of extended-spectrum β-lactamase (ESBL)producing Enterobacterales has increased worldwide in the past 4 decades in community-acquired and nosocomial infections [1].The 30-day mortality rate due to sepsis caused by ESBLproducing Enterobacterales is as high as about 15%, and the choice of antimicrobial agent significantly affects patient outcomes [2].The number of antimicrobial agents susceptible to ESBLproducing Enterobacterales is limited, and ESBL-producing Enterobacterales are resistant to most β-lactam and some nonβ-lactam antibiotics, including fluoroquinolones and aminoglycosides [3].Carbapenems (CPMs) are the antimicrobial agents most commonly used worldwide [4].However, excessive CPM use increases the infection rate of CPM-resistant Enterobacterales [5].It also poses a serious mortality problem.Falagas et al calculated that 26% to 44% of all-cause deaths among patients with Enterobacterales infection were attributable to CPM resistance [6].
Cephamycins were isolated from Streptomyces in 1972.Cephamycins, such as cefmetazole (CMZ) and flomoxef, fall under the category of second-generation cephalosporins, and their distinction from other second-generation cephalosporins lies in the presence of a methoxy group at the seventh position of cephalosporanic acid.They are resistant to hydrolysis by ESBL-producing Enterobacterales [3] and have good activity against ESBL-producing Enterobacterales in vitro [7].Cephamycins are widely available and are often used in Japan, Taiwan, and China [8].However, few studies have compared the effects of cephamycin and CPMs on ESBL-producing infections: most reported a small number of cases, and a few that compared CMZ and CPMs showed patients' baseline characteristics comparable to those in our study [9][10][11][12][13].Furthermore, a few articles comparing flomoxef and CPMs for bacteremia caused by ESBL-producing Enterobacterales described the patients' baseline characteristics, including background disease, but most had small sample sizes [14][15][16].

Open Forum Infectious Diseases
The clinical impact of ESBL genes and sequence types has been recently reported [17][18][19].Importantly, since 2008, there has been global dissemination of a specific clone of Enterobacterales: Escherichia coli sequence type 131 (ST131).This has been associated with the rise in ESBL-producing Enterobacterales [17].Additionally, it is suggested to be linked to increased mortality [18].However, limited studies have examined the clinical effect of cephamycins against bacteremia caused by ESBL-producing Enterobacterales.Therefore, the present study aimed to address this knowledge gap by increasing the sample size and examining a more diverse patient population.Furthermore, we aimed to examine the ESBL gene and sequence type distribution and evaluate how gene and sequence type could influence the comparative study between the CMZ and CPM groups.

METHODS
This retrospective cohort study was conducted at Tenri Hospital, a 715-bed hospital in Nara, Japan.Cases were reviewed for patients who were treated with CMZ or CPMs for bacteremia caused by ESBL-producing Enterobacterales between 1 April 2014 and 31 September 2022.This study followed the STROBE statement (Strengthening the Reporting of Observational Studies in Epidemiology) [20].The Institutional Review Board of Tenri Hospital approved this study protocol (No. 1216), and the requirement for informed consent was waived.
The criteria for screening ESBL-producing Enterobacterales were per the methods of the Clinical and Laboratory Standards Institute and the European Committee on Antimicrobial Susceptibility Testing: minimal inhibitory concentration ≥8 μg/mL for cefpodoxime or ≥2 μg/mL for ceftazidime, aztreonam, cefotaxime, or ceftriaxone for E coli, Klebsiella pneumoniae, and Klebsiella oxytoca; minimal inhibitory concentration ≥2 μg/mL for cefpodoxime, ceftazidime, or cefotaxime for Proteus mirabilis [21].ESBL production was confirmed with a double-disk synergy test with cefotaxime, ceftriaxone, cefepime, and an amoxicillin-clavulanate disk [22].
We included patients who were >18 years of age, had at least 1 episode of monomicrobial bacteremia caused by ESBL-producing Enterobacterales, and had been treated with CMZ or CPMs as definitive therapy within 3 days after antimicrobial susceptibility data were identified.We excluded patients who had ESBL-producing Enterobacterales resistant to either CMZ or CPMs; a history of an allergic reaction to CMZ, CPMs, or other products in the same class; and pregnancy, lactation, or intentions to become pregnant during the study.ESBL-producing Enterobacterales were defined as positive strains in the doubledisk synergy set.The following data were collected: • Patient characteristics: age, sex, height, body weight, department in which patients were admitted (in our hospital, the Department of General Internal Medicine admits patients with renal and collagen disease), and immunocompromised status • Charlson Comorbidity Index, community-acquired or nosocomial infection, origin of bacteremia, causative pathogens, and Pitt bacteremia score (PBS) • The treatment profile outlines the antimicrobials used within 72 hours and the application of drainage therapy We classified patients as immunocompromised if they met any of the following criteria: taking prednisone >2 mg/kg or >20 mg daily for at least 14 days, having received a biological agent in the previous 30 days, having a history of solid organ transplant, undergoing a hematopoietic stem cell transplant within the past year, receiving cancer chemotherapy within the last 6 months, possessing any congenital immunodeficiency, or living with HIV with a CD4 count of <200 cells/μL [23].
The primary outcome measure was 90-day all-cause mortality.Secondary outcomes included 30-day mortality, recurrent episodes of bacteremia, readmission within 30 days, duration of antibiotic use, and adverse events.Furthermore, we identified the resistance genes of ESBL-producing Enterobacterales.Bacterial DNA was purified with the QIAmp DNA Mini Kit (Qiagen).For E coli ST131 typing, ST131 was defined by polymerase chain reaction detection of ST131-specific single-nucleotide polymorphisms in the mdh and gyrB alleles [20].In β-lactamase gene typing, strains were analyzed to determine the presence of ESBL encoded by bla CTX-M-1 , bla CTX-M-2 , and bla CTX-M-9 [24,25].

Statistical Analysis
We compared the baseline characteristics of the patients in the CMZ and CPM groups.In addition, we divided the patients into ST131 and non-ST131 groups and compared them with the CMZ and CPM groups.We excluded missing data and used data sets for which all the variables were available.Patient characteristics were described by median (IQR) for continuous variables and number (percentage) for categorical variables.A Mann-Whitney U test was used to compare continuous variables between the groups, whereas a Fisher exact test was used to compare categorical variables.The condition for statistical significance was defined as P < .05.Statistical analyses were performed with SPSS version 22 (IBM).

Patient Characteristics
A total of 233 patients had positive blood cultures for bacteremia caused by ESBL-producing Enterobacterales, of which 156 were included in the analysis.Among them, 90 (58%) received CMZ therapy while 66 (42%) received CPM.Among the CPM group, meropenem was used in 62 patients, followed by imipenem-cilastatin in 3 and doripenem in 1 (Figure 1).In all cases, the doses of antimicrobials were adjusted by type of drug and renal function.The characteristics of the patients are shown in Table 1.Patients in the CMZ group were significantly older than those in the CPM group (median [IQR], 79 years [71-86] vs 74 years [64-83]; P = .011),and there were significantly more females in the CPM group (48 [52%] vs 16 [24%], P = .010).The severity of the PBS in the CMZ group was significantly lower than that in the CPM group (median [IQR], 0 [0-2] vs 2 [0-2]; P = .042).Moreover, no patients in the CMZ group were admitted to the Department of Hematology, and the CPM group had significantly more patients who received a hematopoietic stem cell transplant in the preceding year or cancer chemotherapy within 6 months.However, there were no significant differences in the number of patients who received prednisone ≥2 mg/kg or ≥20 mg daily for at least 14 days or biological agents in the preceding 30 days; in addition, analysis of the other factors, including the Charlson Comorbidity Index, revealed no statistical differences.

Microbiological Characteristics
The CMZ and CPM groups had antimicrobial resistance genes such as bla CTX-M-1-like (36%, 23%), bla CTX-M-2-like (2%, 3%), and bla CTX-M-9-like (61%, 71%).Patients who died within 90 days harbored bla CTX-M-1-like (n = 3, 4) and bla CTX-M-9-like (n = 3, 6) genes.The analysis of ESBL gene types is presented in Table 3. ST131 clones were detected in 66 patients (73%) in the CMZ group and 43 (65%) in the CPM group.There were no significant differences in the prevalence of ST131 between the groups (P = .293).In the cohort with non-ST131 clones, no patients in the CMZ group and 3 in the CPM group died within 90 days.We conducted a subgroup analysis of the primary and secondary outcomes of ST131 and non-ST131 clones (Table 4).While hospital-acquired infections, immunocompromised hosts, and PBS were significantly higher in the CPM group of ST131 clones (P = .013,P = .029,P = .020,respectively), there was no significant difference in the primary outcome between the CMZ and CPM groups in either cohort.Among the ST131 clones, there were significantly more patients readmitted within 30 days in the CPM group than the CMZ group (P < .001),while non-ST131 clones did not demonstrate significant differences.

DISCUSSION
To the best of our knowledge, our retrospective study examined the largest sample size to date on patients treated with CMZ or CPMs for bacteremia caused by ESBL-producing Enterobacterales.Furthermore, our study included several patients with hematologic tumors and neutropenic fever, which were not included in previous studies.Our results suggest similar efficacies against bacteremia caused by ESBL-producing organisms.Fukuchi et al reported the efficacy of CMZ as a definitive therapy against bacteremia caused by ESBL-producing Enterobacterales as compared with CPM.Their results revealed no significant differences in 90-day mortality rate between the groups (1/26 vs 5/43) [10].Matsumura et al showed that the treatment of patients with  .64 Prosthetic valve/intracardiac implantable device 5 (6) 5 (8) .743 Intracranial catheter 0 (0) 0 (0) Chest tube 0 (0) 1 (2) .423 Percutaneous endoscopic gastrostomy 1 (1) 3 (5) .311 Source bacteremia caused by ESBL-producing E coli with CMZ or flomoxef was similarly effective as treatment with CPM (30-day mortality was 3/59 vs 5/54, respectively) when the patients did not have hematologic malignancy or neutropenia [13].Moreover, the number of the adverse events seems to be as small as that in previous studies [10,13], which might indicate a benefit of the CMZ being with the narrow spectrum.Based on the findings of our study, CMZ emerges as a potential therapeutic option for the management of bacteremia caused by ESBL-producing Enterobacterales in patients with hematologic malignancies and/or neutropenic fever.Patients with ESBL-producing K pneumoniae bloodstream infections have higher mortality than those with ESBL-producing E coli [26,27].CPMs have been recommended for ESBL-producing Klebsiella bacteremia in previous studies [28,29].However, there have been limited data on the effectiveness of CMZ for ESBL-producing K pneumoniae bacteremia.
Two previous reports stated the efficacy of CMZ when compared with CPM.Fukuchi et al reported that the proportion of Klebsiella spp was 6% (4/69; 2 cases of K pneumoniae and 2 cases of K oxytoca) [10], whereas Matsumura et al analyzed the bacteremia caused by E coli only [13].Yet, in our study, the prevalence of Klebsiella spp was 11% (18/158 cases; 17 cases of K pneumoniae and 1 case of K oxytoca).This prevalence is higher than in the previous 2 studies.Moreover, in our study, the mortality rate caused by Klebsiella spp bacteremia was 0% (0/8) in the CMZ group and 40% (4/10) in the CPM group.Our study indicated the possible efficacy of CMZ in treating ESBL-producing Enterobacterales, including Klebsiella spp.
We analyzed the genotypes and sequence types of ESBL-producing Enterobacterales in the present study.CTX M15 β-lactamases are the most common type of ESBLs identified in Europe and some countries in Asia, Africa, North America, South America, and Australia [30].Similar to previous ESBL genotyping studies conducted in Japan [31][32][33], CTX M9 was predominant in our study.When a subanalysis was conducted for each genotype group, there was no notable disparity in primary and secondary outcomes between the CMZ group and the CPM group.This observation aligns with the limited existing studies, which have not consistently reported differences in mortality based on genotype.The prevalence of ESBL-producing ST131 strains has increased rapidly [34].The prevalence rates of ST131 clones in previous studies (62% and 43%) were not significantly different from those in the present study [30,31].The clinical impact of ST131 in patients with bacteremia caused by ESBL-producing Enterobacterales has remained controversial [17], but Wang et al reported that the ST131 clone was associated with a higher 28-day mortality in patients with bacteremia caused by ESBL producers [34].In this study, ST131 clones might have affected the readmission  rate within 30 days, but the presence or absence of ST131 clones did not influence other characteristics and outcomes.The genotype and sequence type results indicated that CMZ was as effective as CPM, regardless of the sequence type.This study had several limitations.First, this was a retrospective study conducted at a single institution.Second, there was a noticeable difference between the groups in the number of patients who were admitted to the Department of Hematology orwho had received chemotherapy, as well as in PBS result and receipt of bone marrow transplant.Third, some patients were treated with empiric antibiotics before definitive therapy.However, CPM was used before CMZ in only 3 patients (3.3%) in the CMZ group.

CONCLUSION
The findings of our retrospective study suggest that CMZ is a well-tolerated alternative to CPM for treating bacteremia caused by ESBL-producing Enterobacterales.More extensive prospective studies in multiple settings and multicenter randomized trials are required to corroborate these findings.

Notes
Acknowledgments.We express our gratitude to the members of the Microbiological Research Centre for their excellent care and assistance.This research was conducted as part of the All-Osaka U Research in "The Nippon  Foundation-Osaka University Infectious Disease Response Project."