Abstract
Background: Induction of antibiotic resistance is associated with increased morbidity and mortality in AmpC β-lactamase producing Enterobacteriaceae. The use of ceftriaxone is controversial for treatment of these organisms due to concerns for inducible resistance. This study was designed to compare treatment failure rates between ceftriaxone and antipseudomonal β-lactam antibiotics when used as definitive therapy for organisms most commonly associated with chromosomal AmpC β-lactamase production. Methods: A retrospective, single-center cohort study was performed enrolling patients hospitalized with monomicrobial Enterobacter, Citrobacter, or Serratia spp. infections. The primary objective compared proportion of treatment failure between groups. All patients received either ceftriaxone or an antipseudomonal β-lactam alone within 24 hours of culture finalization, and with a duration of at least 72 hours for definitive treatment. Treatment failure was defined as either clinical failure (abnormal white blood cell count or temperature on day 7 or 14 post-antibiotics) or microbiologic failure (regrowth of the same organism at same site within 14 or 21 days). Results: Of 192 total patients, treatment failure was observed in 24/71 patients (34%) receiving ceftriaxone and in 42/121 patients (35%) receiving antipseudomonal β-lactam (P = .98). No difference was observed between clinical or microbiologic failure rates between groups. The ceftriaxone group had significantly more patients undergoing treatment for urinary tract infections (51% vs 17%, P < .001), but treatment failure rates remained similar between groups when comparing infections of all other sources. Conclusion: Ceftriaxone has comparable treatment failure rates to antipseudomonal β-lactams for susceptible Enterobacteriaceae infections and may be considered as a therapeutic option. Further, prospective research is needed to validate optimal dosing and application in all sites of infection.
Keywords: AmpC, Enterobacteriaceae, ceftriaxone, piperacillin-tazobactam, cefepime, meropenem, treatment failure
Introduction
A common resistance mechanism among Enterobacteriaceae is β-lactam hydrolysis by β-lactamase enzymes.1,2 The AmpC-mediated β-lactamase is characterized as a serine-active cephalosporinase (Molecular class C, Functional Group 1), and is commonly found in chromosomes of some species of Enterobacteriaceae including Enterobacter, Citrobacter, and Serratia.3 -5 Organisms expressing AmpC can be particularly challenging to treat due to its inducible properties, which promote bacterial survival.6 -9 Production of the β-lactamase is suppressed at baseline; however, enzyme production can increase 10- to 100-fold in the presence of cell wall fragments caused by specific β-lactam antibiotics.8,9 As a result, it is possible that the bacteria appear susceptible in vitro, but develop resistance during a course of therapy.6,7
Enterobacteriaceae infections are regularly documented within health-systems. Estimates of culture-positive Enterobacter spp. and Serratia spp. infections account for 7% and 3.5% of intensive care unit (ICU) infections worldwide, respectively.10,11 β-lactam therapy must be chosen very intentionally in these infections, with particular regard to induction and susceptibility to the β-lactamase. Those antibiotics that are both strong inducers and susceptible to the β-lactamase will likely fail over the course of treatment. As such, AmpC β-lactamase production precludes the use of aminopenicillins, cefazolin, and cephamycins in most cases. 6 The β-lactams least likely to be hydrolyzed include cefepime or the carbapenem class, resulting in widespread use in clinical practice.6 -9,12,13 One area of significant controversy is the place of ceftriaxone in the treatment of infections by potential AmpC β-lactamase-producing organisms without knowledge of true AmpC production. It is unclear whether third-generation cephalosporins are appropriate therapy as they are not strong inducers of this enzyme but they are unstable in the presence of the enzyme.6 -9 Some studies have reported ceftriaxone in vivo resistance rates up to 19%,14,15 while other literature suggests clinical failure rates as low as 5% to 10%.16 -20
As the worldwide threat of bacterial resistance increases, it is critical to optimize the use of available antimicrobials and limit potentially unnecessary broad-spectrum use. It is still unknown if ceftriaxone is effective in the treatment of susceptible infections caused by species of Enterobacter, Citrobacter, and Serratia. This knowledge could provide clinicians confidence in decisions to de-escalate antibiotics for these common nosocomial infections. The objective for this study was to evaluate whether ceftriaxone could exhibit similar treatment failure compared to antipseudomonal β-lactam antibiotics in infections caused by potential AmpC producing organisms.
Methods
Patients and Data Collection
This retrospective cohort study evaluated all adult patients with a culture-positive Enterobacter, Citrobacter, or Serratia spp. infection. Patients hospitalized within a single health-system from 2013 to 2016 were assessed for inclusion. IRB approval was obtained for this project. The primary outcome assessed rates of treatment failure. Secondary outcomes assessed organism colonization and compared ceftriaxone and cefepime subgroups. Patients were included if they had Enterobacter, Citrobacter, or Serratia isolates grown on any microbiologic culture and were treated with either ceftriaxone or an antipseudomonal β-lactam agent as definitive therapy. The original culture had to be susceptible to the definitive antibiotic chosen based on the 2009 Clinical and Laboratory Standards Institute (CLSI) breakpoints which were utilized by the lab in the study time frame. 21 Patients with extended-spectrum β-lactamase producing isolates were enrolled if this stipulation was met. Antipseudomonal β-lactams on formulary included meropenem, piperacillin-tazobactam, and cefepime. Cultures were not assessed for the phenotypic presence of AmpC β-lactamase as this testing was not available at the study institution.
Enrollment was stratified based on organism to include Enterobacter isolates more frequently than Serratia or Citrobacter isolates. A ratio of 3:1:1 was used due to high Enterobacter prevalence and greater AmpC β-lactamase production in this species. A separate patient list for each antibiotic and species was generated and each list was randomized using a random number generator prior to screening. The investigators enrolled all ceftriaxone patients that met inclusion criteria. Patients treated with meropenem, cefepime, and piperacillin-tazobactam were enrolled via stratification that allowed for a subgroup comparison between ceftriaxone and cefepime alone. Patients were excluded if they were prisoners, pregnant, immunocompromised, received double coverage for Gram-negative bacteria at time of culture finalization, or if they had a second infectious organism grown on any culture within the clinical assessment window (i.e., concomitant infection). The study timeline for assessment is displayed in Figure 1. Patients’ location was determined based on their location at the time of the initial culture. Severity of illness was assessed using the Sequential Organ Failure Assessment (SOFA) score. 22 If any SOFA score data points were not monitored by the treating physician for all 7 days of clinical assessment, it was assumed the patient had no organ dysfunction in that category. ICD 9 and 10 codes were used to identify and categorize relevant comorbidities as determined in an a priori fashion.
Figure 1.
Study timeline: the study could be initiated at any point during the patient’s stay. For infections requiring an initial prescription >14 days, the clinical and microbiologic windows were extended to days 14 and 21, respectively.
Definitions
Definitive therapy was defined as the antibiotic that the patient was on within 24 hours of culture finalization or day 4 of treatment, whichever was later. The patient must have been on the study antibiotic for at least 72 hours during the course of therapy. Treatment failure was defined as either clinical failure or microbiologic failure. Clinical failure was based on the presence of an abnormal white blood cell count (≤4000 cells/µL or ≥12 000 cells/µL) or temperature (≤36°C or ≥38°C) in the absence of a second infection. 23 These markers for clinical failure were assessed 7 days after the initial culture was obtained (Figure 1). Microbiologic failure was defined as regrowth of the same organism in the same site at an equal or greater amount as the initial culture. Blood cultures were considered for microbiologic failure if there was any organism regrowth regardless of quantitative or qualitative amount. Microbiologic failure was monitored from days 2 to 14 after the initial culture was drawn. For treatment that had a duration of longer than 14 days, clinical failure was assessed on day 14 and microbiologic failure assessed through day 21.
A distinction was made in the experiment protocol to identify patients with colonization of bacteria but who would not be considered as a treatment failure. To this end, organism colonization was different from microbiologic failure in regards to both time frame of the analysis and the amount of organism growth. Colonization was analyzed up to 30 days after the initial culture was procured and the qualitative or quantitative amount of growth distinguished microbiologic failure from colonization if the time frame was within days 2 to 14.
Immunocompromised patients were those who had an absolute neutrophil count of <1000/µL, received a previous solid-organ transplant, bone marrow transplant within 12 months, or chemotherapy within the past 6 months; were infected with the human immunodeficiency virus; or had a documented congenital immunodeficiency. 12
Type of infection was determined based on the type of culture that was drawn. Quantitative or qualitative analysis was not performed for diagnosis of infection, because all patients were determined to be clinically infected and prescribed antibiotics by the treating physician. Patients who had multiple sources of infection with the same organism were included in the individual analysis of each site they grew the organism (e.g., a patient with a positive urine culture and blood culture would be included in both urinary tract infection [UTI] and bloodstream analysis).
Statistical Analysis
Patients were categorized based on the use of ceftriaxone or antipseudomonal β-lactam antibiotics. Chi-squared or Fisher’s exact tests were used to compare categorical data. Student’s t-test and Mann–Whitney Rank Sum were used to analyze and compare continuous data as appropriate based on Shapiro–Wilk testing for normality. All statistics were calculated using SigmaPlot® 13.0 software (Systat®, San Jose, CA). Subgroup analyses were performed post hoc with the exception of an analysis of the antipseudomonal β-lactam class.
Multivariate logistic regression was performed to identify factors associated with treatment failure. Covariates identified a priori included SOFA score, prior use of β-lactam antibiotics within 14 days, length of antibiotic therapy, ICU status, species of infective organism, and site of infection. To determine which species of organism and which antibiotics would be included in the model, a univariate analysis was performed. Variables with a P-value <.2 could be included in the multivariate logistic regression. The best-fit model with the identified variables was determined using a backward stepwise regression.
To estimate our sample size, a failure rate for antipseudomonal β-lactams of 8% was assumed based on a range of 0% to 16% for drugs in this class reported in previous studies.12,14,15,17 -20 A 20% absolute difference in treatment failure rates was determined to be clinically significant and has been previously described with the use of third generation cephalosporins.14,15 Sixty-seven patients in each group were needed for enrollment to meet 80% power with an alpha level of 0.05.
Results
Description of the Cohort
One hundred and ninety-two patients met study eligibility; 71 received ceftriaxone and 121 received an antipseudomonal β-lactam (Figure 2). The antipseudomonal β-lactam group included 80 patients on cefepime, 27 on piperacillin-tazobactam, and 14 on meropenem. The majority of patients on meropenem screened for enrollment were excluded due to polymicrobial cultures.
Figure 2.
Study enrollment: patients were enrolled into the antibiotic group chosen as their definitive therapy. They may have been started on any antibiotic prior to culture finalization.
Note. TZP = piperacillin-tazobactam.
Other baseline characteristics are displayed in Table 1. The ceftriaxone group was older (median [inter-quartile range]: 66 [51-80] vs 56 [40.5-71] years, P = .01), had higher Sequential Organ Failure Assessment (SOFA) scores (3 [1-6] vs 2 [1-5], P = .04), and was more likely to have comorbid chronic obstructive pulmonary disease (11 [15.5%] vs 4 [3.3%], P = .03). Urinary sources of infection were more prevalent in the ceftriaxone group (36 [50.7%] vs 21 [17.4%], P < .001). Thirty-five patients (49.2%) in the ceftriaxone group were treated empirically with an antipseudomonal β-lactam prior to definitive therapy with ceftriaxone. Total duration of antimicrobial therapy was shorter in the ceftriaxone group (8 days [7-11] vs 9.5 days [7-14.75], P = .049).
Table 1.
Baseline Characteristics as Charted on the Day a Positive Culture for Enterobacter, Citrobacter, or Serratia Was Obtained.
| Ceftriaxone n = 71 | Antipseudomonal β-lactam n = 121 | P -value | |
|---|---|---|---|
| Demographic information: | |||
| Median age, years (IQR) | 66 (51-80) | 56 (40.5-71) | .01 |
| Male, n (%) | 39 (54.9) | 84 (69.4) | .06 |
| BMI, mean kg/m2 ± SD | 28.6 ± 8.9 | 27.8 ± 8.2 | .52 |
| Location, n (%) | |||
| Floor | 26 (36.6) | 56 (46.2) | .19 |
| ICU | 45 (63.3) | 65 (53.7) | |
| Median SOFA (IQR) | 3 (1-6) | 2 (1-5) | .04 |
| β-lactam in prior 14 days, n (%) | 28 (39.4) | 47 (38.8) | .94 |
| Hospital LOS, median days (IQR) | 14 (8-23) | 14 (9-25) | .73 |
| Presenting treatment failure indicators | |||
| Median temperature (C; IQR) | 37.2 (36.7-38.1) | 37.1 (36.7-38) | .72 |
| Median WBC (cells/L; IQR) | 11.3 (7.5-16.9) | 12.1 (8.3-17.4) | .46 |
| Comorbid conditions | |||
| Diabetes, n (%) | 18 (25.3) | 41 (33.8) | .27 |
| COPD, n (%) | 11 (15.4) | 4 (3.3) | .03 |
| Alcoholism, n (%) | 5 (7.0) | 4 (3.3) | .42 |
| Infectious species | |||
| Enterobacter spp. | 40 (56.3) | 69 (57.0) | |
| E. cloaceae | 29 (40.8) | 47 (38.8) | |
| E. aerogenes | 10 (14.1) | 21 (17.3) | |
| Citrobacter spp. | 12 (16.9) | 11 (9.1) | |
| C. freundii | 4 (5.6) | 4 (3.3) | |
| Serratia spp. | 19 (26.8) | 41 (33.9) | |
| S. marcescens | 18 (25.3) | 39 (32.2) | |
| Infection type | |||
| Urinary tract, n (%) | 36 (50.7) | 21 (17.4) | <.001 |
| Respiratory tract, n (%) | 27 (38.0) | 41 (33.9) | .67 |
| Bloodstream, n (%) | 7 (9.9) | 29 (24.0) | .03 |
| SSTI, n (%) | 2 (2.8) | 23 (19.0) | .003 |
| Osteomyelitis, n (%) | 1 (1.4) | 20 (16.5) | .003 |
| Multiple sources of infection, n (%) | 6 (8.5) | 21 (18.2) | .103 |
| Duration of treatment (days) median [IQR] | 8 [7-11] | 9.5 [7-14.75] | .049 |
Note. ICU = intensive care unit; SOFA = Sequential Organ Failure Assessment; LOS = length of stay; C = Celsius; WBC = white blood cell; COPD = chronic obstructive pulmonary disease; SSTI = skin and soft tissue infection.
In-vitro resistance patterns are displayed in Table 2. Ten (5.2%) organisms were susceptible to cefazolin utilizing the 2009 CLSI breakpoints while 176 of all cultures (91.6%) were susceptible to ceftriaxone. 20 Of the antipseudomonal β-lactam class, resistance was only observed to piperacillin-tazobactam in 17 isolates (9.9%). A post hoc analysis compared the organisms’ minimum inhibitory concentration (MIC) to the updated 2010 CLSI Standards. Seventy out of 71 isolates in the ceftriaxone group exhibited a ceftriaxone MIC < 1 mcg/mL with one isolate having an MIC of 8 μg/mL. All isolates in the cefepime group (n = 80) had a cefepime MIC < 2 μg/mL.
Table 2.
Organism Sensitivity Data by Species.
| Cefazolin susceptible*
MIC < 4 μg/mL, n (%) |
Ceftriaxone susceptible MIC < 8 μg/mL, n (%) |
TZP susceptible MIC < 16 μg/mL, n (%) |
FEP/MEM susceptible MIC < 2 μg/mL, n (%) |
|
|---|---|---|---|---|
| Enterobacter, n = 109 | 0 (0) | 95 (87) | 93 (85) | 108 (99) |
| E. cloacae, n = 76 | 0 (0) | 65 (86) | 63 (83) | 75 (99) ǂ |
| E. aerogenes, n = 31 | 0 (0) | 28 (90) | 28 (90) | 31 (100) |
| Other E. spp., n = 2 | 0 (0) | 2 (100) | 2 (100) | 2 (100) |
| Citrobacter, n = 23 | 10 (43) | 21 (91) | 22 (96) | 23 (100) |
| C. freundii, n = 8 | 0 (0) | 7 (88) | 7 (88) | 8 (100) |
| Other C. spp., n = 15 | 10 (67) | 14 (93) | 15 (100) | 15 (100) |
| Serratia, n = 60 | 0 (0) | 60 (100) | 60 (100) | 60 (100) |
| S. marcescens, n = 57 | 0 (0) | 57 (100) | 57 (100) | 5 (100) |
| Other S. spp., n = 3 | 0 (0) | 3 (100) | 3 (100) | 3 (100) |
| Total | 10/191 | 176/191 | 175/191 | 190/191 |
Units are number of isolates grown from microbiologic laboratory.
Note. MIC = minimum inhibitory concentration; TZP = piperacillin-tazobactam; FEP = cefepime; MEM = meropenem.
Definitions taken from 2009 CLSI breakpoints and the institution laboratory could not test MICs for cefazolin below 4 mcg/mL.
Only one isolate exhibited an MIC > 2 for cefepime. It’s MIC of 4 mcg/mL was considered susceptible by the 2009 breakpoints but would be intermediate according to 2010 CLSI breakpoints. This isolate was treated with meropenem and included in the meropenem group.
Primary and Secondary Outcomes
Treatment failure did not differ between ceftriaxone and antipseudomonal β-lactam groups (24/71 [33.8%] vs 42/121 [34.7%], respectively, P = .98). Clinical failure was the most common cause of treatment failure in both ceftriaxone and antipseudomonal β-lactam groups. Microbiologic failure was higher in the antipseudomonal β-lactam group, but this was not statistically significant. No significant differences in clinical failure or microbiologic failure were observed between individual agents. Mortality occurred in 3/71 ceftriaxone patients compared to 10/121 in the antipseudomonal β-lactam group (P = .25).
Of those patients included, 33/192 [17.2%] had microbiological colonization with at least one culture drawn that identified the same organism within 30 days. Of these patients, eight had received ceftriaxone [11.3%] and 25 received antipseudomonal β-lactams [20.7%] (P = .142). The highest rate of colonization was observed with piperacillin-tazobactam (9/27 [33.3%]); this was statistically significant compared to all other antibiotics (P = .04).
A multivariate logistic regression was performed to identify independent predictors of treatment failure. Variables identified for inclusion based on univariate analysis were use of cefepime and infections secondary to Serratia spp. UTI was forced into the model post hoc to control for confounds in baseline characteristics. Infections due to Serratia spp. (odds ratio [95% confidence interval]: 2.48 [1.25-4.92], P = .01) and ICU status (2.25 [1.05-4.80], P = .04) were independent predictors of treatment failure. No individual antibiotic was an independent predictor of treatment failure.
Subgroup Analyses
A priori subgroup analyses of antipseudomonal β-lactam antibiotics were performed for the primary and secondary endpoints. Treatment failure occurred in: cefepime-23/80 [29%], meropenem-8/14 [57%], piperacillin-tazobactam-11/27 [41%], (P = .09).There were no statistically significant differences among any of the agents for clinical failure, microbiologic failure, or colonization. Significantly fewer patients were treated for UTI in the piperacillin-tazobactam group (23% vs 21% vs 0%, respectively, P = .03). When UTIs were excluded, there were still no significant differences between cefepime, meropenem, and piperacillin-tazobactam in regards to treatment failure (31% vs 64% vs 41%, respectively, P = .101).
A post hoc subgroup analysis was conducted comparing outcomes between cephalosporins. Patients that received ceftriaxone had higher SOFA scores (3 [1-6] vs cefepime, 2 [0-4.75], P = .01) and rate of urinary source infection (36/71 [50.7%] vs 23/80 [28.8%], P < .001). No significant differences were observed between ceftriaxone and cefepime in regards to treatment failure (24/71 [33.8%] vs 23/80 [28.8%], P = .62) or organism colonization (8/71 [11.3%] vs 12/80 [15.0%], P = .66).
There were no significant differences among treatment failure rates between ceftriaxone and antipseudomonal β-lactam antibiotics between different sources of infection (Table 3). However, there was significantly less colonization in patients treated with ceftriaxone for urinary source infection (0/35 [0%] vs 4/21 [19.0%], P = .02). Treatment failure and colonization rates were similar among the ceftriaxone and the antipseudomonal β-lactam groups controlling for UTI. Twenty-seven of 71 (38.0%) patients that received ceftriaxone were treated for a pulmonary source of infection, similar to the rates of pulmonary infection in the antipseudomonal group (41/121 [33.9%]). Rates of treatment failure were comparable (12/27 [44.4%] vs 20/41 [48.8%], P = .92) for pulmonary source infections. Having multiple sources of infection did not impact treatment failure (13/28 [46.4%] vs 53/164 [32.3%], P = .22).
Table 3.
Treatment Failure Assessed by Infection Type and Treatment Arm. a
| Average days of therapy | Treatment failure | P-value b | Colonization, 30 days | P-value | |
|---|---|---|---|---|---|
| Urinary tract infection | |||||
| Ceftriaxone, n = 36, n (%) | 8.4 | 10 (28) | .99 | 0 (0) | .02 |
| APBL, n = 21, n (%) | 9.5 | 5 (24) | 4 (19) | ||
| Respiratory infection | |||||
| Ceftriaxone, n = 27, n (%) | 10.2 | 12 (44) | .92 | 6 (22) | .93 |
| APBL, n = 41, n (%) | 8.9 | 20 (49) | 10 (24) | ||
| Bloodstream | |||||
| Ceftriaxone, n = 7, n (%) | 13.6 | 4 (57) | .68 | 1 (14) | 1.00 |
| APBL, n = 29, n (%) | 16.7 | 12 (41) | 7 (24) | ||
| Non-UTI | |||||
| Ceftriaxone, n = 35, n (%) | 12.1 | 14 (40) | .961 | 8 (23) | .99 |
| APBL, n = 100, n (%) | 14.8 | 37 (37) | 21 (21) | ||
Note. UTI = urinary tract infection.
SSTI (n = 25), osteomyelitis (n = 21), and intra-abdominal infection (n = 9) removed from analysis due to small number of ceftriaxone patients (2, 1, 3, respectively).
P-value reflects comparison of treatment failure.
Discussion
Clinical and microbiologic failure were similar between ceftriaxone and the antipseudomonal β-lactams for the treatment of potential AmpC producing organisms. This study represents one of the largest cohorts treated with ceftriaxone as definitive therapy for Enterobacter, Citrobacter, and Serratia spp. infection after concerns arose for inducible resistance. 14 The results of this study question the generalization that all infections by these organisms cannot be effectively treated with ceftriaxone. Its tolerability and ease of administration highlight further benefits of expanding its place for treatment. One study suggests that, while resistance to ceftriaxone may develop more rapidly with Enterobacter spp., resistance to cefepime that develops during therapy may confer consequences even beyond the β-lactam class. 24 In addition, each additional day of antipseudomonal therapy has been associated with increased risk of future resistant infections. 25 Given the ecological impact of unnecessary broad-spectrum antimicrobial use, there is a need to confirm the place of ceftriaxone for the treatment of these prevalent nosocomial infections.
There was a disparity of the sites of infection between the two treatment groups. However, the success of ceftriaxone was not limited to UTI alone. Some researchers have suggested that all non-urinary Enterobacter, Citrobacter, Serratia, Providencia, and Morganella spp. be labeled resistant to third generation cephalosporins, 26 but the results of the study suggest ceftriaxone use may be effective for infections of other sources. Thirty-four patients in the ceftriaxone group (47.8%) were treated for either respiratory or bloodstream infections and demonstrated no difference in treatment failure or colonization (Table 3). In addition, urinary source was not identified as an independent predictor for treatment success in this study. These findings should be interpreted with caution, as the largest numerical disparity in treatment success was observed in patients with bacteremia. This is consistent with published data suggesting the highest rates of emergent resistance to third generation cephalosporins are observed in patients with Enterobacter bacteremia.14,15
Dosing is an important aspect of beta-lactam therapy as time above the MIC (fT > MIC) must be at least 40% to 70% to correlate to efficacy. 27 Patient populations included in this study, including critically ill patients, can challenge the ability for standard dose beta-lactams to maintain optimal fT > MIC. Strategies such as prolonged infusions and higher dosing may be employed to overcome these challenges.27,28 All ceftriaxone doses used in the study were infused over 30 minutes. The median daily dose [IQR] for all ceftriaxone patients was 2 g/day [1-4], and when UTI patients were removed the median dose increased to 4 g/day (2 g intravenously every 12 hours). Critically ill patients exhibiting augmented renal clearance or hypoalbuminemia may require higher daily doses of ceftriaxone to achieve treatment success. 29 Patients treated with antipseudomonal beta-lactams received a median daily dose of 4 g cefepime [2-6] and 10.125 g for piperacillin-tazobactam [9-10.125] as extended infusions. Antimicrobial dosing was not associated with treatment failure in this study.
No phenotypic characterization of AmpC production was performed at any point during the study. This highlights an important distinction between stable derepression and induction of the AmpC β-lactamase alone. Stable derepression occurs when regulation of the AmpC β-lactamase is removed, creating large quantities of the enzyme itself. 9 Neither ceftriaxone nor piperacillin-tazobactam would likely yield success in this scenario as both antibiotics would be hydrolyzed. It is possible that this was the case in earlier literature that cited high rates of ceftriaxone failure. Induction of the enzyme alone may not yield clinically significant amounts of the β-lactamase without stable derepression. This could mean that in organisms with in vitro sensitivity to third-generation cephalosporins, which is estimated to be as high as 50% to 98% in these Enterobacteriaceae, treatment success can be possible.30,31
An interesting finding were the high rates of treatment failure and colonization among patients who received piperacillin-tazobactam for definitive treatment. Careful interpretation is warranted when evaluating this subgroup. First, piperacillin-tazobactam has been used with success in previous studies when used for definitive therapy against these organisms.32,33 It is also important to emphasize that colonization was not a reinfection rate (e.g., microbiologic failure), but instead highlights patients with a decreased eradication of the organism or a site that remains colonized with bacteria. Since piperacillin-tazobactam was not used for the treatment of any urinary source infections in this study, increased colonization rates may be a spurious finding requiring further research. Further comparison between antipseudomonal β-lactams is warranted for selection between these agents.
A similarly cautious approach should be considered when evaluating the high rates of treatment failure for both meropenem patients and patients with Serratia spp. infections. A majority of meropenem patients (10, 71%) were located in the ICU at time of enrollment with multiple factors that could influence a systemic response. Patients with a Serratia spp. infection had a higher median SOFA score at baseline (4 [1.25-6] vs 2 [1-5], P < .001) compared to all other patients. It is reasonable to correlate treatment failure with higher severity of illness in these populations.
A strength of this study is the large number of patients on ceftriaxone that met inclusion criteria for definitive treatment. In addition, the large population of cefepime patients enrolled allowed for a comparison between cephalosporin agents along with the antipseudomonal β-lactam group. The analysis of organism colonization also lends insight into the possible ramifications of expansion of ceftriaxone use beyond treatment failure alone. Finally, data from this study identify possible dosing optimization opportunities for ceftriaxone use in these organisms, though further research is required in this area.
There are several limitations to this study, starting with its retrospective and single health-system design. The disparity of infection types that each antibiotic was treating may also lead to heterogeneity. Nevertheless, analysis of the failure rates were similar among individual infection types. Another limitation is the high treatment failure rate. The definition used was likely conservative as it identified more than 30% of patients in each group as failing therapy. This should not be interpreted as a lack of efficacy of β-lactams in general but rather related back to a poor resolution of fever or leukocytosis within a 7-day period. Most patients in this study failed due to inadequate resolution of their leukocytosis. While the results were sensitive to identify most cases of treatment failure, this was likely at the expense of specificity. The analysis of treatment failure and dosing of antibiotics may be confounded by selection bias, as patients treated for UTIs received lower daily doses of antibiotics than those treated for other sources of infection. Finally, the low rates of skin/soft tissue infection in the ceftriaxone group limit the extrapolation of the results beyond urinary, respiratory, or bloodstream source infections.
In conclusion, ceftriaxone may be considered in the treatment of potential AmpC β-lactamase-producing organisms. It is prudent for prescribers to consider optimization of dose when treating infections of a non-urinary source. However, it is unclear if ceftriaxone is effective in the treatment of bacteremia due to the low number and seemingly high failure rate in this subset of patients. Further research is warranted to determine the most appropriate duration and dose of therapy, as well as clinical utility of characterizing stable derepression of AmpC β-lactamase.
Acknowledgments
The authors would like to express gratitude to Riane Ghamrawi, PharmD, BCPS, and G. Christopher Wood, PharmD, BCCCP, FCCM, FCCP for their critical review and helpful commentary on the manuscript.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs: David M. Peters 
https://orcid.org/0000-0002-3313-7359
Siyun Liao
https://orcid.org/0000-0002-4433-0445
References
- 1. Livermore DM. Bacterial resistance: origins, epidemiology, and impact. Clin Infect Dis. 2003;36(suppl 1):S11-S23. doi: 10.1086/344654. [DOI] [PubMed] [Google Scholar]
- 2. Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev. 2010;74(3):417-433. doi: 10.1128/MMBR.00016-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bush K. 2018. Past and present perspectives on β-lactamases. Antimicrob Agents Chemother. 2018;62(10). doi: 10.1128/AAC.01076-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ambler RP. The structure of beta-lactamases. Philos Trans R Soc Lond B Biol Sci. 1980;289(1036):321-331. doi: 10.1098/rstb.1980.0049. [DOI] [PubMed] [Google Scholar]
- 5. Bush K, Jacoby GA. Updated functional classification of beta-lactamases. Antimicrob Agents Chemother. 2010;54(3):969-976. doi: 10.1128/AAC.01009-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Tamma PD, Doi Y, Bonomo RA, Johnson JK, Simner PJ. A primer on AmpC beta-lactamases: necessary knowledge for an increasingly multidrug-resistant world. Clin Infect Dis. 2019;69(8):1446-1455. doi: 10.1093/cid/ciz173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Jacoby GA. AmpC β-lactamases. Clin Microbiol Rev. 2009;22(1):161-182. doi: 10.1128/CMR.00036-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. MacDougall C. Beyond susceptible and resistant, part I: treatment of infections due to gram-negative organisms with inducible β-lactamases. J Pediatr Pharmacol Ther. 2011;16(1):23-30. [PMC free article] [PubMed] [Google Scholar]
- 9. Jones RN. Important and emerging β-lactamase-mediated resistances in hospital-based pathogens: the Amp C enzymes. Diagn Microbiol Infect Dis. 1998;31(3):461-466. [DOI] [PubMed] [Google Scholar]
- 10. Vincent JL, Rello J, Marshall J, et al. International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009;302(21):2323-2329. doi: 10.1001/jama.2009.1754. [DOI] [PubMed] [Google Scholar]
- 11. Trilla A. Epidemiology of nosocomial infections in adult intensive care units. Intensive Care Med. 1994;20(S4):S1-S4. doi: 10.1007/bf01745243. [DOI] [PubMed] [Google Scholar]
- 12. Tamma PD, Girdwood SC, Gopaul R, et al. The use of cefepime for treating AmpC β-lactamase-producing Enterobacteriaceae. Clin Infect Dis. 2013;57(6):781-788. doi: 10.1093/cid/cit395. [DOI] [PubMed] [Google Scholar]
- 13. Rodríguez-Baño J, Gutiérrez-Gutiérrez B, Machuca I, Pascual A. Treatment of infections caused by extended-spectrum-beta-lactamase-, AmpC-, and carbapenemase-producing Enterobacteriaceae. Clin Microbiol Rev. 2018;31(2). doi: 10.1128/CMR.00079-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Chow JW, Fine MJ, Shlaes DM, et al. Enterobacter bacteremia: clinical features and emergence of antibiotic resistance during therapy. Ann Intern Med. 1991;115(8):585-590. doi: 10.7326/0003-4819-115-8-585. [DOI] [PubMed] [Google Scholar]
- 15. Kaye KS, Cosgrove S, Harris A, Eliopoulos GM, Carmeli Y. Risk factors for emergence of resistance to broad-spectrum cephalosporins among Enterobacter spp. Antimicrob Agents Chemother. 2001;45(9):2628-2630. doi: 10.1128/aac.45.9.2628-2630.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Goldstein FW. Cephalosporinase induction and cephalosporin resistance: a longstanding misinterpretation. Clin Microbiol Infect. 2002;8(12):823-825. [DOI] [PubMed] [Google Scholar]
- 17. Siedner MJ, Galar A, Guzmán-Suarez BB, et al. Cefepime vs other antibacterial agents for the treatment of Enterobacter species bacteremia. Clin Infect Dis. 2014;58(11):1554-1563. doi: 10.1093/cid/ciu182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Choi SH, Lee JE, Park SJ, et al. Emergence of antibiotic resistance during therapy for infections caused by Enterobacteriaceae producing AmpC beta-lactamase: implications for antibiotic use. Antimicrob Agents Chemother. 2008;52(3):995-1000. doi: 10.1128/AAC.01083-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Fish DN, Piscitelli SC, Danziger LH. Development of resistance during antimicrobial therapy: a review of antibiotic classes and patient characteristics in 173 studies. Pharmacotherapy. 1995;15(3):279-291. [PubMed] [Google Scholar]
- 20. Milatovic D, Braveny I. Development of resistance during antibiotic therapy. Eur J Clin Microbiol. 1987;6(3):234-244. [DOI] [PubMed] [Google Scholar]
- 21. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. 19th ed. CLSI Supplement M100. Wayne, PA: Clinical and Laboratory Standards Institute; 2009. [Google Scholar]
- 22. Vincent J-L, Moreno R, Takala J, et al. The SOFA (sepsis-related organ failure assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707-710. doi: 10.1007/bf01709751. [DOI] [PubMed] [Google Scholar]
- 23. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest. 1992;101(6):1644-1655. [DOI] [PubMed] [Google Scholar]
- 24. Fung-Tomc JC, Gradelski E, Huczko E, Dougherty TJ, Kessler RE, Bonner DP. Differences in the resistant variants of Enterobacter cloacae selected by extended-spectrum cephalosporins. Antimicrob Agents Chemother. 1996;40(5):1289-1293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Teshome BF, Vouri SM, Hampton N, Kollef MH, Micek ST. Duration of exposure to antipseudomonal β-lactam antibiotics in the critically ill and development of new resistance. Pharmacotherapy. 2019;39(3):261-270. doi: 10.1002/phar.2201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Livermore DM, Brown DF, Quinn JP, Carmeli Y, Paterson DL, Yu VL. Should third-generation cephalosporins be avoided against AmpC-inducible Enterobacteriaceae? Clin Microbiol Infect. 2004;10(1):84-85. [DOI] [PubMed] [Google Scholar]
- 27. Shah S, Barton G, Fischer A. Pharmacokinetic considerations and dosing strategies of antibiotics in the critically ill patient. J Intensive Care Soc. 2015;16(2):147-153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Roberts JA, Lipman J. Pharmacokinetic issues for antibiotics in the critically ill patient. Crit Care Med. 2009;37(3):840-851. [DOI] [PubMed] [Google Scholar]
- 29. Ollivier J, Carrié C, D’houdain N, et al. Are standard dosing regimens of ceftriaxone adapted for critically ill patients with augmented creatinine clearance? Antimicrob Agents Chemother. 2019;63(3):1-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Pfaller MA, Jones RN, Marshall SA, et al. Inducible Amp C β-lactamase producing gram-negative bacilli from blood stream infections: frequency, antimicrobial susceptibility, and molecular epidemiology in a national surveillance program (SCOPE). Diagn Microbiol Infect Dis. 1997;28(4):211-219. [DOI] [PubMed] [Google Scholar]
- 31. Park KH, Chong YP, Kim SH, et al. Impact of revised broad-spectrum cephalosporin clinical and laboratory standards institute breakpoints on susceptibility in Enterobacteriaceae producing AmpC β-lactamase. Infect Chemother. 2017;49(1):62-67. doi: 10.3947/ic.2017.49.1.62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Cheng L, Nelson BC, Mehta M, et al. Piperacillin-tazobactam versus other antibacterial agents for treatment of bloodstream infections due to AmpC β-lactamase-producing Enterobacteriaceae. Antimicrob Agents Chemother. 2017; 61(6):1-10. doi: 10.1128/AAC.00276-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Harris PNA, Tambyah PA, Lye DC, et al. 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. 2018;320(10):984-994. doi: 10.1001/jama.2018.12163. [DOI] [PMC free article] [PubMed] [Google Scholar]