Ceftazidime/avibactam (CZA) is an essential treatment option for managing infections caused by multidrug-resistant (MDR) Gram-negative (G−) bacteria, including complicated intra-abdominal infections (cIAI), complicated urinary tract infections (cUTI), hospital-acquired pneumonia (HAP), ventilator-associated pneumonia (VAP), and associated bacteremia (European Medicines Agency, 2024). Approved by the Food and Drug Administration (FDA) in 2015 and the European Medicines Agency (EMA) in 2016, CZA effectively targets key pathogens like Enterobacteriaceae (e.g., Klebsiella) and Pseudomonas aeruginosa (European Medicines Agency, 2024), which have become a significant global healthcare challenge due to rising carbapenem resistance. Studies have highlighted the genetic diversity of carbapenem-resistant Klebsiella pneumoniae (CRKP), with subspecies producing New Delhi metallo-β-lactamase (NDM) and oxacillinase-48 (OXA-48)-like carbapenemases, with varying lineages [e.g., sequence type (ST)-14, ST-16, ST-231] (Ceron et al., 2023; Garcia-Gonzalez et al., 2023; Chen et al., 2023). The Mediterranean region is particularly noted for endemic strains of Klebsiella that produce OXA-48 oxacillinase (Bonomo et al., 2018). P. aeruginosa resistance to carbapenems is driven by the production of MBLs (VIM, IMP, NDM, SPM), mutations in porin channels and the overexpression of efflux pumps (Poole, 2011).

Recent European data underscore the growing threat. For K. pneumoniae, data from the European Antimicrobial Resistance Surveillance Network (EARS-Net) for 2023 shows a significantly increasing trend of population-weighted mean percentage of carbapenem (imipenem or meropenem) resistant isolates from 10.4% in 2019 to 13.3% in 2023 (European Centre for Disease Prevention and Control, 2023a). Significantly increasing national trends in carbapenem resistance percentages were observed in invasive isolates of K. pneumoniae during the period 2019–2023 for Croatia and many other European countries (European Centre for Disease Prevention and Control, 2023b). Estimated incidence of bloodstream infections (BSIs) of CRKP for the EU/EEA increased from 2.52 in 2019 to 3.97 per 100,000 population in 2023, which constitutes an increase of 57.5% (European Centre for Disease Prevention and Control, 2023a). Estimated incidence of BSIs of carbapenem-resistant P. aeruginosa for the EU/EEA increased from 1.73 in 2019 to 2.01 per 100,000 population in 2023, which constitutes an increase of 16.2% (European Centre for Disease Prevention and Control, 2023a). Carbapenem-resistant P. aeruginosa is listed by World Health Organisation (WHO) as a pathogen of high priority that requires research and development of new antibiotics (World Health Organization, 2024).

The rising prevalence of MDR pathogens, combined with limited treatment options and high associated mortality, highlights the urgent need for optimized therapeutic strategies and infection control. Effective antibiotic regimens play a crucial role in combating resistant bacterial strains, whereas antibiotic resistance poses a significant threat to public health (Batchelder et al., 2023). Strategies like antibiotic cycling, evolutionary steering, and combination therapies have been proposed to slow down resistance development and enhance treatment outcomes (Ghazi and El Nekidy, 2023). Additionally, the rediscovery and optimization of dosing of older antibiotics have shown promise in treating infections caused by resistant bacterial strains, albeit with considerations for potential adverse effects (Ali et al., 2022). Avibactam is a potent β-lactamase inhibitor, making CZA active against 94%–100% of Klebsiella Pneumoniae-carbapenemase (KPC) and OXA-48-producing pathogens from the Enterobacteriaceae family (Tamma et al., 2023; Kazmierczak et al., 2018). However, resistance to CZA has also been observed (Cui et al., 2023), necessitating the exploration of dosing strategies that optimize pharmacokinetic/pharmacodynamic (PK/PD) targets.

Antimicrobial activity of CZA is primarily defined by the percent time of the free-drug concentration above the CZA minimum inhibitory concentration (MIC) for ceftazidime, and the percent time of the free-drug concentration above a threshold concentration (Ct) for avibactam (Das et al., 2019), over the dosing interval. Built on previous findings, including 50% fT > MIC as a well-established target for the efficacy of ceftazidime alone (Muller et al., 2013), MIC values for CZA ranging from 4–8 mg/L for P.aeruginosa (Nichols et al., 2016; Flamm et al., 2014) and 0.5–4 mg/L for Enterobacteriaceae (Flamm et al., 2014; De Jonge et al., 2016), and threshold concentration for avibactam of at least 1 mg/L to be effective for both Enterobacteriaceae and P.aeruginosa (Coleman et al., 2014; Berkhout et al., 2016), the joint PK/PD target for CZA was established (50% fT > 8 mg/L for ceftazidime and 50% fT > Ct of 1 mg/L for avibactam) (Das et al., 2019). The selected target was used in population pharmacokinetic modelling to determine a dosing regimen of 2 g/0.5 g every 8 h via a 2-h intravenous infusion, for all the aforementioned indications. The recommended regimen, with dose adjustment in patients with creatinine clearance (ClCr) < 50 mL/min (Das et al., 2019), has been evaluated in five randomised Phase III clinical trials (RECLAIM 1 and 2, RECLAIM 3, REPRISE, RECAPTURE, REPROVE) (Mazuski et al., 2016; Wagenlehner et al., 2016; Carmeli et al., 2016; Qin et al., 2017; Torres et al., 2018), that all confirmed its efficacy and safety. A recent post hoc pooled analysis of these trials further reinforced the clinical effectiveness and safety of CZA in patients with infections caused by ESBLs, AmpC and serine carbapenemase-producing G-pathogens (Torres et al., 2023). However, its suitability for critically ill patients in intensive care units (ICUs) with altered PK profiles remains unclear. Early post-marketing evaluations highlighted the urgent need for PK investigations in this high-risk population to optimize dosing, minimize risk of underexposure, and prevent the emergence of resistance (Falcone and Paterson, 2016). Currently approved intermittent dosing (ID) regimens for β-lactams generally have more than an 80% probability of achieving the PK/PD target, but data for ICU patients are very limited. It is believed that a target of 100% fT ≥ 4×MIC, and according to some sources, even 100% fT ≥ 5×MIC in patients with severe infections in the ICU would contribute to clinical efficacy (Al Shaer et al., 2020; Roberts et al., 2014). It is also important to note that the PK/PD target necessary to prevent resistance is often higher than the one needed for clinical efficacy. In critically ill patients, factors such as increased capillary permeability, volume redistribution, large fluid infusions, low serum albumin, and interventions like mechanical ventilation can alter β-lactam pharmacokinetics, notably the volume of distribution (Vd) and clearance (Cl). Renal clearance is often reduced but may also be elevated, especially in sepsis, further complicating drug distribution, particularly in septic shock where microcirculatory impairment reduce drug concentration at infection sites. Hydrophilic drugs like ceftazidime are particularly susceptible to fluctuating plasma levels, making standard dosing regimens, derived from healthy volunteers’ PK data, potentially inadequate for critically ill patients. The effectiveness of β-lactams can be difficult to measure early on, especially within the first 24–48 h of therapy, and therapeutic drug monitoring (TDM) is not widely available. This can obscure whether poor outcomes are due to illness severity or underdosing in ICUs.

Studies reveal significant pharmacokinetic variability in ceftazidime among ICU patients with P. aeruginosa infections and severe sepsis (Rondanelli et al., 1986; Young et al., 1997; Gómez et al., 1999). Notably, substantial interindividual differences in half-life (t1/2) and trough levels of ceftazidime reflect variations in Vd and Cl (Rondanelli et al., 1986; Young et al., 1997). Even in pediatric ICU patients, recent population PK data showed high variability in ceftazidime Cl, with body weight and eGFR identified as key covariates (Li et al., 2024). Given that β-lactams’ bactericidal effect depends on the time that drug concentration exceeds MIC, extending infusion duration may counter pharmacokinetic shifts, keeping plasma levels above MIC for a longer time. It has been confirmed that the continuous infusion (CI) of ceftazidime in critically ill patients with HAP maintains clinical efficacy, optimizes the PD profile, ensures concentrations above the MIC for a longer period (Cousson et al., 2005; Nicolau et al., 2001), and allows for the use of a smaller total dose of the drug (Nicolau et al., 2001; Angus et al., 2000) compared to ID. Greater clinical efficacy of ceftazidime in CI compared to ID has been demonstrated in patients with VAP caused by G-pathogens (Lorente et al., 2007). A pharmacokinetic study by El Haj et al. (2024) demonstrated that CI also provided superior anti-biofilm activity in vitro compared to ID in both susceptible and resistant P. aeruginosa strains, highlighting the PD benefits of sustained drug exposure. A pharmacoeconomic comparison of continuous versus intermittent ceftazidime use showed that CI is a cost-effective alternative to ID (McNabb et al., 2001).

Growing evidence indicates that critically ill patients are also at an increased risk of altered PK of CZA and that optimal PK/PD exposures may not be achieved with standard dosing regimens (Bakdach et al., 2022; Cojutti et al., 2024). Suboptimal exposure to CZA can contribute to the development of resistance in Enterobacteriaceae pathogens, particularly in critically ill patients (Adembri et al., 2020). This was demonstrated in a case report of a critically ill patient with KPC-producing K. pneumoniae bacteremia where suboptimal CZA exposure led to pathogen mutation and the development of resistance (Gaibani et al., 2021). Gatti and Pea suggested that implementing modified dosing strategies such as extending infusion duration or increasing doses, together with TDM, could be the most effective approach to optimizing PK/PD in critically ill patients (Gatti and Pea, 2021a). This strategy may also aid in eradicating difficult-to-treat (DTR) G-pathogens, reduce the risk of microbiological treatment failure, and prevent unnecessary use of combination therapy (Gatti and Pea, 2021a; Gatti et al., 2023a; Gatti et al., 2023b; Situm et al., 2024). Pharmacokinetic analyses further support that CI can overcome dosing limitations, particularly for pathogens requiring aggressive PK/PD targets (Cojutti et al., 2024). Continuous infusion could also reduce the total daily dose, as lower doses than those recommended in the standard dosing regimens may still achieve the desired PK/PD target (Gatti et al., 2023b; Fresan et al., 2023). A review of real-world data from 2021 found that prolonged infusion (3 h) of CZA was negatively associated with mortality; however, only 23.7% of patients were hospitalized in the ICU (Gatti and Pea, 2021b).

Despite growing evidence supporting CI of CZA in critically ill patients, current studies remain limited by small sample sizes, observational designs, and variability in dosing regimens. Consequently, robust, high-quality evidence is still lacking to definitively establish the clinical superiority of CI over ID in this population. The primary hypothesis of this study is that CI of CZA improves microbiological success compared to ID in critically ill ICU patients with severe infections due to Klebsiella pneumoniae OXA-48 or Pseudomonas aeruginosa. The central research question is whether CI of CZA improves microbiological outcomes compared to ID in this high-risk population.

The ZAVICONT trial addresses an urgent need to optimize β-lactam therapy in critically ill patients with severe infections caused by MDR G-bacteria, specifically K. pneumoniae OXA-48 and carbapenem-resistant P. aeruginosa. Due to their limited treatment options, rapid dissemination in hospital settings, and high associated mortality rates, these pathogens pose a significant global threat to public health.

Ceftazidime/avibactam has emerged as a crucial treatment for these infections. However, despite its relatively recent clinical introduction, resistance to CZA has already been reported, particularly when suboptimal drug exposure permits pathogens to evolve resistance mechanisms (Cui et al., 2023; Adembri et al., 2020; Gaibani et al., 2021). This is of a special concern in ICU settings, where ID regimens often fail to achieve optimal PK/PD targets due to altered drug PK (Bakdach et al., 2022; Cojutti et al., 2024). The ZAVICONT trial aims to determine whether CI can achieve superior PK/PD outcomes and improve clinical efficacy, leveraging the time-dependent nature of β-lactam activity. This trial builds on previous findings suggesting that CI could be the most effective approach to optimize CZA PK/PD targets, facilitate eradication of DTR G-pathogens, and reduce the risk of microbiological treatment failure while minimizing total daily drug doses (Cojutti et al., 2024; Gatti and Pea, 2021a; Gatti et al., 2023a; Gatti et al., 2023b; Situm et al., 2024). To date, this is one of the first randomized trials specifically evaluating CI versus ID for CZA in critically ill patients (Figure 4), providing much-needed evidence to guide clinical practice in ICU settings.

The primary outcome of the trial, microbiological success, is a crucial endpoint for evaluating pathogen eradication and is directly influenced by achieving adequate PK/PD targets. Secondary outcomes, including clinical success, length of ICU and hospital stay, and all-cause 28-day mortality, will provide a comprehensive assessment of the clinical and economic benefits of CI over ID. The incorporation of the ceftazidime-to-MIC ratio (C/MIC) as a PK/PD marker underscores the trial’s focus on precision medicine, addressing variability in drug exposure among critically ill patients. The trial’s design incorporates practical considerations, such as ensuring the stability of CZA during extended infusions and addressing logistical challenges in ICU environments.

The ZAVICONT trial, while robust, has several limitations that were carefully considered during its design to ensure feasibility and applicability. First, its single-center design may limit the generalizability of findings to other settings with different patient populations, microbiological profiles, or ICU practices. This design choice was made to maintain strict control over patient selection, drug administration, and data collection, ensuring consistency and minimizing variability that could arise from differences in protocols across multiple centers. However, the trial is being conducted in a high-volume tertiary ICU with clinical practices and patient profiles representative of other advanced European ICUs. The study was designed around a clearly defined clinical question, standardized intervention protocols, and commonly encountered pathogens, which enhances its external relevance and provides a solid methodological foundation for future multicenter validation. Second, the lack of blinding could introduce potential bias, particularly in the assessment of subjective outcomes such as clinical success. The decision to not include blinding was based on both logistical and ethical considerations. Administering a placebo alongside the active drug would significantly increase the workload for ICU staff and introduce unnecessary fluid burden in critically ill cardiac patients, many of whom are already at risk of volume overload. To minimize bias, both data collection and outcome assessment will be performed by personnel who are blinded to treatment allocation. Third, the study measures only ceftazidime plasma concentrations, without avibactam quantification. This was due to unavailability of a validated avibactam assay in our country at the time of study initiation. Developing new method would have significantly delayed study initiation. While this approach is consistent with several prior PK studies, we acknowledge that optimal TDM of CZA should ideally assess both components, with reference to publication by Gatti et al. (2024). We plan to implement avibactam monitoring in future studies.

We acknowledge that potential confounders, such as concomitant antimicrobial therapies, variability in ICU practices, and patient-level factors including comorbidities, source control interventions, and organ support could influence study outcomes. To mitigate their impact, eligibility criteria were narrowly defined, and randomization was applied to ensure balance between groups. Moreover, standardized treatment protocols and microbiological procedures were implemented. Relevant concomitant therapies will be documented and considered in sensitivity analyses where applicable. Additionally, we recognize that evolving antimicrobial resistance patterns during the study period may influence microbiological outcomes. This represents an important contextual factor in the interpretation of results, particularly when considering the emergence of new resistance mechanisms to CZA or shifts in local epidemiology.

These design choices ensure that the ZAVICONT trial remains both feasible and highly relevant, with findings that are directly translatable to routine clinical practice while paving the way for future research. Importantly, these limitations do not compromise the study’s internal validity or its ability to address the primary research question, reinforcing the trial’s robustness.

Future directions include prospective multicenter trials to validate these findings across different ICU settings and patient populations. Additional analyses may explore pharmacoeconomic implications, impact on resistance development, and subgroup outcomes based on specific comorbidities or infection sources. By addressing these broader clinical questions, the ZAVICONT trial can serve as a foundation for more personalized and effective β-lactam therapy in critically ill patients.

This study will be conducted in accordance with all applicable guidelines for ethical research, including the principles outlined in the Declaration of Helsinki, the International Conference on Harmonisation (ICH) guidelines for Good Clinical Practice (GCP), the Healthcare Act of the Republic of Croatia (NN 100/18), and the Patients' Rights Act of the Republic of Croatia (NN 169/04). The study protocol, participant information sheet, and informed consent form were approved by the Ethics Committee of UHC Zagreb. Prior to enrollment, all participants will provide written informed consent. Given the critically ill nature of the participant population, the use of deferred consent will be applied when the participant is temporarily unable to provide consent, in accordance with national regulations and Article 30 of the Declaration of Helsinki. In such cases, informed consent will be obtained from a legally authorized representative. If no representative is immediately available and if inclusion in the study cannot be delayed, the patient may be enrolled without prior consent, provided that this is explicitly justified in the protocol and the procedure has received prior Ethics Committee approval. Consent to remain in the study will be sought as soon as possible from the patient or their legal representative. Additional safeguards will be implemented to ensure that participants and/or their representatives fully understand the potential risks and benefits of participation. Participant data confidentiality will be strictly maintained by assigning each participant a unique numerical code for data storage in electronic databases. Access to sensitive data will be restricted to authorized personnel only. A risk-adapted safety monitoring strategy has been introduced and is detailed in the Methods section.

The findings of the ZAVICONT trial will be shared widely to maximize their impact on clinical practice and scientific knowledge. Results will be published in peer-reviewed journals focused on infectious diseases, critical care, and pharmacology to ensure broad audience within relevant fields. Additionally, the study team will present key results at both national and international conferences.