Temocillin versus meropenem for the targeted treatment of bacteraemia due to third-generation cephalosporin-resistant Enterobacterales (ASTARTÉ): protocol for a randomised, pragmatic trial.

INTRODUCTION: Alternatives to carbapenems are needed in the treatment of third-generation cephalosporin-resistant Enterobacterales (3GCR-E). Temocillin is a suitable candidate, but comparative randomised studies are lacking. The objective is to investigate if temocillin is non-inferior to carbapenems in the targeted treatment of bacteraemia due to 3GCR-E. METHODS AND ANALYSIS: Multicentre, open-label, randomised, controlled, pragmatic phase 3 trial. Patients with bacteraemia due to 3GCR-E will be randomised to receive intravenously temocillin (2 g three times a day) or carbapenem (meropenem 1 g three times a day or ertapenem 1 g once daily). The primary endpoint will be clinical success 7-10 days after end of treatment with no recurrence or death at day 28. Adverse events will be collected; serum levels of temocillin will be investigated in a subset of patients. For a 10% non-inferiority margin, 334 patients will be included (167 in each study arm). For the primary analysis, the absolute difference with one-sided 95% CI in the proportion of patients reaching the primary endpoint will be compared in the modified intention-to-treat population. ETHICS AND DISSEMINATION: The study started after approval of the Spanish Regulatory Agency and the reference institutional review board. Data will be published in peer-reviewed journals. TRIAL REGISTRATION NUMBER: NCT04478721. © Author(s) (or their employer(s)) 2021. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

The design of this randomised study has limitations, including the open design, the heterogeneity of oral alternatives for switching and the exclusion of patients with septic shock.

One of the strengths is its pragmatic design which we hope will allow the appropriate representation of patients with the target infections.

The multicentre participation and the short time limit to recruit the patients once the bacteraemia is diagnosed are other strengths of this study.

Introduction

Infections due to antimicrobial-resistant pathogens are recognised as a worldwide public health problem. The problem is especially severe among Gram-negative bacteria. In fact, third-generation cephalosporin-resistant Enterobacterales (3GCR-E), either caused by extended-spectrum β-lactamases (ESBLs) or high production of AmpC enzymes, were the leading cause of invasive infections (estimated, 365 000) and attributable deaths (estimated, 12 700) among antibiotic-resistant bacteria in the European economic area in 2015.1 Also, 3GCR-E have a very important impact in the use of antibiotics; a very important increase in carbapenems consumption (the drugs of choice for invasive infections due to 3GCR-E) has followed these bacteria spread2 and it is contributing to the subsequent explosive spread of carbapenems resistance.

Therefore, alternative treatments for 3GCR-E are desperately needed. One of the alternatives is piperacillin-tazobactam, but its efficacy compared with carbapenems is controversial.3 4 The cephamycins may be active against ESBL-producers, but not against AmpC-producers, and their efficacy is doubtful.5 Fosfomycin and aminoglycosides may be an empirical option in some cases, but they are only useful in urinary tract infection (UTI).6 7 Finally, new β-lactams (ceftolozane-tazobactam, ceftazidime-avibactam) might be reserved for other multidrug-resistant pathogens.8

Temocillin is a β-lactam drug which is stable against ESBLs and AmpC enzymes, and therefore is active against a high proportion of 3GCR-E.9 This drug is only approved in a few countries for the treatment of septicaemia, urinary tract and lower respiratory tract infections when susceptible Gram-negative bacilli are suspected or confirmed (standard dosing, 2 g two times a day intravenously; for severe infections, 2 g three times a day is recommended). The pharmacokinetic and pharmacodynamic properties of temocillin have recently been reviewed.10

The objective of this article is to describe the hypothesis, objectives, design, variables and procedures for a pragmatic randomised controlled trial comparing the efficacy of temocillin and meropenem in bacteraemic infections caused by 3GCR-E. To the best of our knowledge, no randomised trials have been published with temocillin in these infections.

Methods and analysis

Hypothesis and objectives of the trial

The hypothesis of the study is that temocillin is non-inferior to carbapenems for the targeted treatment of bacteraemia due to 3GCR-E. The primary objective of the trial is to demonstrate the non-inferiority of temocillin in terms of efficacy and safety. Secondary objectives include providing specific comparative data about the efficacy and safety of temocillin and carbapenems in subgroups of patients (eg, different sources of bacteraemia, elderly, renal insufficiency and other underlying conditions), providing data about the pharmacokinetics and pharmacodynamics of temocillin, and about the distribution of the minimum inhibitory concentrations (MIC) of temocillin according to the mechanisms of resistance.

Trial design, sites and study period

ASTARTÉ is a multicentre, open-label, randomised, controlled, pragmatic phase 3 trial. A 36-month recruitment period is planned. The study is coordinated from Hospital Universitario Virgen Macarena (HUVM) under the auspices of the Spanish Network for Research in Infectious Diseases, the Andalusian Network for Clinical Research in Infectious Diseases and the Spanish Clinical Research Network (SCReN). The trial is funded by Instituto de Salud Carlos III, Spanish Ministry of Science and Innovation; the drug is kindly provided by Belpharma SA (Luxemburg) under an agreement with states that Belpharma SA will be informed about the results of the study but cannot influence the analyses or publication of the results. Thirty-one public Spanish hospitals will participate.

Selection and enrolment

Hospitalised adult patients (≥18 years) with monomicrobial bacteraemia due to Enterobacterales (including Escherichia coli, Klebsiella spp, Proteus spp, Morganella spp, Salmonella spp, Enterobacter spp, Serratia spp, Providencia spp and Citrobacter spp) showing resistance to ceftriaxone, cefotaxime and/or ceftazidime and susceptibility to temocillin and carbapenems are eligible. Inclusion and exclusion criteria are detailed in table 1. Participation in the study is voluntary and patients can withdraw from the study at any time. Subjects will be withdrawn from the study if they experience a major protocol violation, in case of clinical failure or according to safety criteria. Patients will be randomised once inclusion and exclusion criteria are checked (therefore, once the isolate is known to be susceptible to study drug), and informed consent is signed; randomisation must be performed in <96 hours after the blood cultures were obtained and in <48 hours of the availability of susceptibility results.

Table 1

Inclusion and exclusion criteria for participating in ASTARTÉ trial

Inclusion criteria

Exclusion criteria

Adult hospitalised patients with monomicrobial bacteraemia due to Enterobacterales. The micro-organism is resistant to cefotaxime, ceftriaxone (MIC >2 mg/L) and/or ceftazidime (MIC >4 mg/L). The micro-organism is susceptible to temocillin (MIC ≤8 mg/L) and meropenem (MIC ≤2 mg/L). Duration of intravenous treatment is planned to be at least 4 days. The patient signed informed consent.

Age <18 years. Pregnancy or breast feeding. Patients under palliative care. Allergy to beta-lactam drugs. Polymicrobial bacteraemia. Meningitis. Infections typically needing >14 days of therapy (eg, endocarditis, prosthetic joint infection, vascular graft infection, empyema, chronic prostatitis). Severe neutropenia (<500 cells/µL). Septic shock at recruitment. Empirical treatment with an in vitro active drug for >96 hours after initial blood culture extraction.

MIC, minimum inhibitory concentrations.

Randomisation

Candidates will be detected from the daily review of positive blood cultures. Patients with all inclusion criteria but with some exclusion criteria will be considered screening failures. Those signing informed consent will be randomised centrally using an online automatic system with a 1:1 randomisation. Randomisation will be stratified according to empirical treatment (in vitro active or not) and suspected source (urinary tract or other) in order to assure that these variables will be balanced between the study arms. The automatic randomisation system is integrated in the electronic case report form (e-CRF) of the study.

Interventions and study treatment

Patients will be allocated to one of the following arms: Arm A (experimental group), in which patients will receive 2 g of temocillin intravenously every 8 hours in 30–40 min infusion; and Arm B (control group), in which patients will receive 1 g of meropenem intravenously every 8 hours in 15–30 min infusion. Ertapenem 1 g per day can be used instead of meropenem except in patients with sepsis, if MIC ≤0.5 mg/L.

Dosing will be adjusted in patients with renal insufficiency according to official labels (table 2). Duration of intravenous therapy will be decided by the treating physician, but should be at least four full days; then, patients can be switched to an oral regimen if the infection is controlled, the source of infection has been drained/removed/solved, the patient tolerates the oral route and the isolate is susceptible to appropriate drugs as follows: ciprofloxacin 500 mg every 12 hours; trimethoprim-sulfamethoxazole 160–800 mg every 12 hours and amoxicillin-clavulanic acid, 875–125 mg every 8 hours. Recommended duration of total active therapy is 7–14 days. In monomicrobial bacteraemia in which a polymicrobial infection is suspected (eg, intra-abdominal infection), addition of metronidazole (only in patients assigned to temocillin), vancomycin or linezolid (for resistant Gram-positive organisms) or an antifungal agent are allowed.

Table 2

Dose adjustment of study drugs according to renal function

Creatinine clearance (mL/min)	Dose	Frequency
Temocillin (intravenous)
30–60	1 g	Every 12 hours
10–30	1 g	Every 24 hours
<10	500 mg–1 g	Every 24–48 hours
Meropenem (intravenous)
26–50	1 g	Every 12 hours
10–25	500 mg	Every 12 hours
<10	500 mg	Every 24 hours
Ertapenem (intravenous)
<30	1 g	Every 24 hours
<30	Not recommended
Ciprofloxacin (oral)
>60	500 mg	Every 12 hours
30–60	250–500 mg	Every 12 hours
<30	250–500 mg	Every 24 hours
Amoxicillin-clavulanic acid (oral)
10–30	500/150 mg	Every 12 hours
<10	500/125 mg	Every 24 hours
Trimethoprim-sulfamethoxazole (oral)
>30	160/800 mg	Every 12 hours
15–30	80/400 mg	Every 12 hours
<15	Not recommended	Not recommended

Concomitant treatment with any other systemic antibiotic with intrinsic activity against isolated enterobacteria from blood culture is not permitted. The use of one of these antibiotics during the phase of treatment will be deemed as failure and a withdrawal criterion. There are no absolute contraindications for the use of any other drugs during the study.

Since temocillin is not approved in Spain, Belpharma SA will ship the vials to the Pharmacy Service at HUVM, where they will be relabelled and delivered to the sites. The drug traceability will be ensured. The other study drugs are officially approved in Spain, and they will be used through the normal provision of each Hospital Pharmacy at every participating site. The lot number, expiration dates and the number of vials used will be recorded.

Study endpoints

The primary endpoint will be a clinical success in the modified intention-to-treat (mITT) population (see below), and includes all of the following: (1) clinical success at test of cure (TOC); (2) survival at day 28; (3) no need to stop or change the assigned drug because of an adverse event, perceived failure during treatment or occurrence of a superimposed infection; (4) no need to prolong therapy beyond 14 days and (5) no recurrence until day 28. The TOC will be performed 7–10 days after the last day of antibiotic therapy. Clinical success is defined as resolution of the new signs or symptoms related to the infection.

To control potential investigator’s bias, the outcome will be checked through: (1) collection of objective clinical data at day 0 and TOC, including temperature, blood pressure, respiratory and heart rates, Glasgow score and specific examination signs and (2) calculation of the SOFA score (Sequential Organ Failure Assessment) on all visits. A blinded investigator will assess their concordance with the outcome classification provided by the local investigator. Secondary endpoints are shown in table 3.

Table 3

Endpoints in ASTARTÉ

Endpoint	Description	Time of evaluation
Primary endpoint
Clinical success	Proportion of patients with all of the following: (1) clinical cure at TOC (see below); (2) alive at day 28; (3) no need to stop the study drug because of adverse event, failure or intercurrent infection; (4) no need to prolong treatment after 14 days and (5) no recurrence of the infection at day 28	Test of cure (7–10 days after end of treatment)
Secondary endpoints
Clinical cure	Proportion of patients showing resolution of the new signs/symptoms related to the infection	Test of cure (7–10 days after end of treatment)
Mortality	Proportion of dead patients	Day 28
Length of hospital stay	Average time from randomisation to hospital discharge	Hospital discharge
Adverse events	Proportion of patients with any adverse event from first dose of study drug; also of severe adverse events (standard definition)	Day 28
Development of resistence	Proportion of patients with new isolation of the causative micro-organism in follow-up cultures showing resistant to temocillin or meropenem	Day 28
Recurrence	Proportion of patients with development of signs/symptoms of the infection caused by the same micro-organism after clinical response has been reached	Day 28
Reinfection	Proportion of patients with occurrence of a new infection caused by a different micro-organism	Day 28
Change in SOFA score (descriptive endpoint)	Average change in SOFA score	All follow-up visits
Temocillin serum concentrations (descriptive endpoint; only one site)	Distribution of temocillin serum concentration	See text
Temocillin MIC according to mechanisms of resistance (descriptive endpoint)	Distribution of temocillin MIC according to the mechanisms of resistance to cephalosporins	See text

MIC, minimum inhibitory concentrations; SOFA, Sequential Organ Failure Assessment; TOC, test of cure.

Follow-up of participants

Patients will be followed until day 28; all visits and procedures to be performed at each one are specified in table 4. The day of blood culture is considered ‘Day 0’ of the study. After discharge, patients will be provided the means to attend the face-to-face visits.

Table 4

Schedule of follow-up visits and procedures

Assessment	Day 0	Visit 1 selection(Day 1)	Visit 2(Day 3)	Visit 3End of treatment(Days 7–14),	Visit 4Test of cure(14–24)	Visit 5Day 28 (±5)	Unscheduled visit
Randomisation		X
Informed consent		X
Inclusion/exclusion criteria		X
Pregnancy test		X
Clinical history		X	X	X	X	X	X
Physical examination		X	X	X	X	X	X
SOFA scale		X	X	X	X		X
Haematology/chemistry		X	X	X	X		X
Coagulation		X					X
Urinalysis		X			X		X
Blood culture	X		X	X			X
Concomitant medication		X	X	X		X	X
Dispensing control		X	X	X
Adverse events			X	X	X	X	X

*Day 0 is the day of blood culture. Day 1 must not exceed 96 hours after blood culture was drawn.

†If applicable.

‡The visit can be done by telephone if patient is not hospitalised. In this case, physical examination, blood extraction and SOFA scale are not needed.

§If applicable.

¶If it has not been realised in the previous 72 hours.

**Only in urinary tract infection.

††Only if fever persists since visit 2. Repeat it in 48 hours if positive.

‡‡Between 7 and 14 days from the start of the antibiotic treatment.

§§Day 28 (±5 days) since randomisation.

¶¶7–10 days after the end of the treatment.

SOFA, Sequential Organ Failure Assessment.

Microbiological procedures

Blood samples will be performed using standard clinical practice. Blood cultures and bacterial identification will be performed at local laboratories using standard microbiological procedures; the isolates will be preserved and sent to HUVM. Temocillin susceptibility will be studied in 3GCR-E at local laboratories by gradient strips; those with MIC value >8 mg/L will be considered resistant according to British Society of Antimicrobial Chemotherapy breakpoint for susceptibility. Susceptibility to meropenem and other drugs will be studied using routine protocols and interpreted according to the European Committee on Antimicrobial Susceptibility Testing recommendations. Identification and susceptibility in all isolates to all study drugs will be checked later at HUVM using Matrix-Assisted Laser Desorption Ionization Time-of-Flight Mass Spectrometry (MALDI-TOF) and broth microdilution methods, respectively.

Pharmacokinetic and pharmacodynamic studies

Free temocillin plasma levels will be determined on days 1 and 3 in the first 20 patients allocated to the temocillin recruited at HUVM. Blood samples will be obtained 1, 4, 6 and 8 hours after the administration of temocillin; free temocillin plasma concentrations will be measured using HLPC-DAD.11 The method will be validated according to the FDA Bioanalytical Method Validation Guidance for Industry.12 For the population pharmacokinetic analysis, one-compartment and two-compartment linear models will be fitted to the temocillin plasma concentrations-time data. Covariate model building will be performed using sequential assessment of biologically plausible clinical parameters. Monte Carlo models will be built the calculate the probability of target attainment (PTA) ≥50% of the time over the MIC for different MIC values (PTAs for >50% fT>MIC) and simulated dosing schemes (2 g of temocillin administered in 30 min and in 4 hours, every 8 or 12 hours). Dose adjustments will be simulated in patients with decreased renal clearance.

Sample size

We estimated an 85% success rate with meropenem and with temocillin. In order to reject the null hypothesis with 80% power and a 5% one-sided significance level for a 10% non-inferiority margin with a 1:1 assignment, with 5% of missing patients, a total of 167 patients in each study arm are needed (total, 334 patients).

Statistical analysis

For the primary analysis, the absolute difference in the proportion of patients reaching the primary endpoint in the two study arms will be compared in the mITT population, which includes all randomised patients who received at least one dose of the study drug, but in which those incorrectly included or randomised will be excluded. The one-sided 95% CI for the difference will be calculated.

As secondary analyses, all secondary endpoint will be analysed in the mITT population, in the per-protocol population (those receiving at least 4 days of the study drugs) and in the clinically evaluable population (patients evaluated at TOC). Absolute difference with 95% CI will be calculated for categorical endpoints, and Mann-Whitney test for length of hospital stay. The primary endpoint will also be analysed in the following subgroups: according to the source of bloodstream infection (BSI); age groups patients with cancer; mechanism of resistance to third-generation cephalosporins; species of Enterobacterales; temocillin MIC <4 versus 4–8 mg/L; appropriate vs inappropriate empirical therapy; nosocomial versus non-nosocomial episodes and INCREMENT score <7 or ≥8. Analysis considering the site effect will also be performed by using a random effects model. Finally, multivariate analysis will be performed in order to control the potential effect of variables other than randomised antibiotic therapy on the primary outcome using logistic regression and on mortality by Cox regression. Key outcome determinants including age, Charlson, delay in administering an active drug, specific sources, micro-organism, Pitt score, SOFA and renal insufficiency will be considered for inclusion in the models, and will be selected using a stepwise backward process; the variable study arm will be forced in the models.

Safety and adverse event reporting

Pharmacovigilance activities are delegated from the sponsor to the Clinical Trials Unit of University Hospital Virgen del Rocío (CTU-HUVR). Follow-up of adverse events (AE) will be done according to standard procedures and the European Medicine Agency regulation; all potential AE will be recorded and classified according to severity and potential relation with the trial drugs. Any adverse event must be recorded in the e-CRF and all serious AE will be notified in less than 24 hours to CTU-HUVR. The CTU-HUVR personnel are responsible for the reception, recording and resolution of queries and for the identification of any serious unexpected adverse event (SUSAR). SUSAR will be evaluated to communicate them in less than 15 days to Regulatory Authorities, Ethics Committees and Investigators. Safety annual reports will be reported to regulatory Authorities and Ethics Committees by these personnel. A safety monitor from the CTU-HUVR, will coordinate the activities in collaboration with the SCReN.

Data and safety monitoring

Data collection activities will be assessed by an individual responsible of the CTU-HUVR, in contact with the investigators for the revision and verification of data according to a monitoring plan. Subject data will be anonymised and collected using the e-CRF.

An external independent Data Safety Monitoring Board (DSMB) formed by three expert members not participating as investigators in the project will be selected. Interim analyses will be performed after the first 50 and after the first 150 first patients are recruited. Reports with recommendations from the DSMB will be obtained for both interim analyses. A conditional power ≤20% calculated using Mehta and Pocock method after the inclusion of the first 150 patients will be considered low enough to recommend termination of the trial on the basis of futility. Detailed description of rules for decision-making from the committee will be agreed at the time of the agreement of the independent members.

Ethics and dissemination

Ethics

An approved informed consent (version 1.2, dated 6 May 2020) form must be signed before any study specific procedures is performed. The study is approved by the Spanish Regulatory Agency and by the Hospitales Universitarios Virgen Macarena and Virgen del Rocío Ethic Committee. The trial will be carried out according to the principles of the Declaration of Helsinki and the legal Royal Decree RD 1090/2015 applicable in Spain for the performance of clinical trials and European Regulation (EU) n° 536/2014 for all the EU countries.

The results of the study will be submitted for publication to a scientific journal following the Consolidated Standards of Reporting Trials recommendations.

Patient and public involvement

Patient/public involvement will not be involved in the design, or conduct, or reporting, or dissemination plans of our research.

Discussion

Temocillin, because of its in vitro activity, is a potential alternative to carbapenems for the treatment of infections caused by 3GCR-E and might help to reduce the consumption of these drugs.13–15 As comparative clinical data for temocillin is scarce, ASTARTÉ is expected to provide evidence for the use of this drug in the setting of BSI due to 3GCR-E, which represents a substantial proportion of all BSI caused by Enterobacterales.

Because carbapenems are highly efficacious for the treatment of bacteraemia due to 3GCR-E and the objective is to find alternatives which can help in reducing their use, a non-inferiority approach is proposed. The use of an alternative drug might have additional benefits for the patient or the population by reducing the selective pressure of carbapenems for multidrug-resistant organisms. A superiority trial could be done by using a composite primary outcome including colonisation and/or superinfection by multidrug resistant bacteria, but a very high sample size would be needed, making it unfeasible for an investigator-driven clinical trial with public funds.

It is well known that classical randomised clinical trials (RCT) may not be adequately adapted to daily practice; they are frequently performed in selected sites with highly experienced investigators and selected participants, and the population included might not be representative of most patients to whom the results would be extrapolated, so overestimation of benefits and underestimation of harms can be present for special populations normally not included in RCT. This led to the idea that more pragmatic trials showing the real-world effectiveness of the intervention in broader patient groups, are required.16 This may be particularly important in the evaluation of antibiotics as the outcome of the infection do not only depends on the treatment itself but on features of the patients, the source and severity of the infection, the micro-organism and different aspect of the clinical management (source control, support therapy). Therefore, ASTARTÉ was designed as a pragmatic trial trying to mimic clinical practice.

The inclusion of patients with bacteraemia was decided because this is a frequent situation, in which the aetiology is perfectly identified and for which the predictors for outcome have been well studied, also by our group17; the problem of bacteraemia is that it includes different sources of infection, but the experience in previous trials indicates that this can be adequately controlled in the analysis.4 The use of a composite primary endpoint was decided to include both a very relevant and hard variable such as mortality plus clinical success as recommended in a consensus document for trials in bacteraemia.18

Meropenem as comparator was chosen because carbapenems are considered the drugs of choice for invasive infections caused by ESBL producers.19 The use of ertapenem (1 g per day) is accepted except in patients with sepsis if MIC ≤0.5 mg/L.20 In order to approach standard clinical practice, switching to oral therapy when possible is included in the protocol. First option to oral therapy continuation is ciprofloxacin. Trimethoprim-sulfamethoxazole can be used as second option, only in UTI. Third approved option, in case of allergy or resistance to previous treatment described, is amoxicillin-clavulanic acid.

The expected impact of the study is a change in clinical practice allowing temocillin to be used in many patients and contributing to a reduction in the consumption of carbapenems highly needed in the actual situation of resistances.

Strengths and limitations of this study

The design of this randomised study has limitations, including the open design, the heterogeneity of oral alternatives for switching, and the exclusion of patients with septic shock. Some strengths are its pragmatic design which we hope will allow the appropriate representation of patients with the target infections, the multicentre participation and the short time limit to recruit the patients once the bacteraemia is diagnosed.

Funding for the study communicated on November 2019, available for study expenses in January 2020.

Authorisation from the Spanish Regulatory Authority obtained on 9 September 2020.

Approval for the Ethic Committee for the 32 sites included obtained on 5 May 2020.

First patient inclusion for the study occurred on December 2020.