Ceftolozane/tazobactam (C/T) is the combination of ceftolozane, a fifth-generation cephalosporin, with the β-lactamase inhibitor tazobactam. Due to its ability to evade key resistance mechanisms of P. aeruginosa, C/T exhibits potent activity against multidrug-resistant (MDR) and extensively drug-resistant (XDR) P. aeruginosa, including carbapenem-resistant P. aeruginosa (CRPA) [8]. Additionally, it is partially effective against extended-spectrum β-lactamase (ESBL)-producing Enterobacterales, showing a preserved susceptibility in ∼85% of isolates [9, 10]. However, it lacks efficacy against carbapenem-resistant Enterobacterales (CRE). C/T present no efficacy against MDRGNB producing carbapenemases, however, it could retains some sensibility against CRE with other resistance mechanism such as porin-mutation or increase membrane efflux [11, 12]. C/T is approved for the treatment of complicated urinary tract infections (cUTI), complicated intra-abdominal infections (cIAI), and ventilator-associated bacterial pneumonia (VABP). According to both IDSA and ESCMID guidelines, C/T is considered the agent of choice for treating infections caused by difficult-to-treat (DTR) P. aeruginosa [5, 7]. While the IDSA guidelines place C/T on par with CZA and I-R as first-line options, the ESCMID guidelines identify it as the preferred first-line agent. This indication was set as a consequence of the CACTUS study, which has shown a C/T superiority in the treatment of DTR- P. aeruginosa pneumonia, compared to CZA (63% versus 51% in clinical success) [13]. Moreover, CZA has been reported to be associated with higher rates of resistance development; thus, in DTR-P. aeruginosa isolates susceptible to both C/T and CZA, C/T may represent a more appropriate therapeutic option in order to reduce the antibiotic selective pressure [14].

For patients with normal renal function, the approved dosage is 1.5 g administered in 1 h every 8 h for cUTI and cIAI, and 3 g administered in 1 h every 8 h for VABP.

A 2023 study on 163 P. aeruginosa isolates obtained from patients with cystic fibrosis (CF) and LuTx, reported that 81.6% of them were susceptible to C/T [15]. Among MDR and XDR isolates, 88.3% and 28.1%, respectively, were susceptible to C/T. Similarly, Pfaller et al. analysed the susceptibility of 17,315 MDRGNB isolates [16]. They found that P. aeruginosa susceptibility was similar between patients >65 years and immunocompromised hosts (ICH), but notably lower in Intensive care unit (ICU) patients: 96.5% vs. 99.1%/99.2% (ICH/>65 years, respectively) in samples from the US, and 80.1% vs. 93.4%/92.5% in samples from Europe.

Clinical evidence on the use of C/T in LUTR remains limited, summarised in Table 2. In a recent French prospective cohort study involving 63 CF patients, of whom 19% (12/63) were LUTR, with P. aeruginosa lower respiratory tract infections (LRTI) treated with C/T, 89.3% of the strains were susceptible to C/T. The median treatment duration with C/T was 15 days, and clinical improvement was observed in 88.9% of patients [17].

Another multicentre retrospective cohort study of 69 ICH patients, of whom 68% with a history of solid organ transplant (SOT), assessed the outcomes of C/T use in different infections due to MDR P. aeruginosa. The most frequent infection sites were the LRTI (57%). The mean length of C/T therapy was 13 ± 10.8 days. The all-cause 30-day mortality rate among the entire cohort was 19% (13/69), while clinical cure was achieved in 68% (47/69) of patients. This rate was higher in patients with LRTI infections who received 3-g regimens compared to those who received 1.5-g regimens (75% vs. 30%) [18].

Haidal et al. conducted a retrospective study of 21 patients, among whom 7 were LUTR, treated with C/T for MDR P. aeruginosa infections. Most patients (18/21, 86%) had LRTI. The 30-day all-cause and attributable mortality were 10% (2/21) and 5% (1/21), respectively, and the C/T failure rate was 29% (6/21). Resistance to C/T emerged in three patients (14%), primarily associated with de novo mutations. Overexpression and mutations of AmpC were identified as potential mechanisms underlying this resistance [19].

Data regarding the concomitant use of C/T and immunosuppressive agents in SOT recipients remain scarce. Ceftolozane is unlikely to cause clinically relevant drug–drug interactions. Conversely, tazobactam is a substrate of organic anion transporters 1 and 3, and coadministration of inhibitors of these transporters may elevate tazobactam plasma concentrations, warranting cautious monitoring. However, as shown by real-world data, C/T is generally well-tolerated, and the most frequently reported AEs are nausea, vomiting, and diarrhoea.

Clinical data regarding the use of C/T in the LuTx setting are still scarce. C/T is considered the first-line agent for MDR P. aeruginosa, which, in the LuTx setting, is among the most frequently isolated pathogens [23, 24]. In this context, the use of C/T may represent a valuable therapeutic option.

Ceftazidime/avibactam (CZA) combines the third-generation anti-pseudomonal cephalosporin ceftazidime with the novel non-β-lactam BLI avibactam, restoring ceftazidime’s in vitro activity against Ambler class A, class C, and specific class D β-lactamases [25]. However, it remains ineffective against metallo-β-lactamase (MBLs). The primary function of this agent is to treat CRE. To address infections caused by MBL-producing bacteria, CZA is co-administered with aztreonam, taking advantage of their synergistic activity [26]. CZA is currently approved for the treatment of cIAI, UTI, and HBAP/VBAP.

For patients with normal renal function, the recommended dosage is 2.5 g administered in 2 h every 8 h.

The most significant data regarding the use of CZA in LuTx are described in Table 3.

A Chinese retrospective observational study on 15 LUTR, investigating the use of CZA in infections caused by CRPA, reported 14-day and 30-day mortality rates of 6.7% and 13.3%, respectively. Moreover, clinical and microbiological cure rates after CZA therapy were 53.3% and 60% [27].

A similar retrospective study conducted on 10 LUTR treated with CZA for carbapenem-resistant K. pneumoniae (CRKP) and CRPA infections showed 30-day and 90-day survival rates of 100% and 90%, respectively. However, recurrent CRKP and CRPA infection did occur in 50% of patients [31].

An international, retrospective cohort study evaluated the efficacy of CZA compared with best available therapy (BAT) in a cohort of 149 SOT recipients with BSI caused by CRKP. LUTR accounted for 4% of the overall SOT population, and among the 83 patients treated with CZA, two were LUTR. Treatment with CZA was associated with a significantly higher rate of clinical success at day 14 compared to BAT (80.7% vs. 60.6%). A similar pattern was observed for clinical success at day 30, with statistically significant differences favouring CZA [27].

Notably, CZA treatment was associated with improved survival outcomes in the CAVICOR study, which represents the largest cohort to date investigating the impact of CZA on mortality in infections caused by CRE. However, only 45 out of 339 patients (13.2%) included in the analysis were SOT recipients, and no stratification by type of transplant was provided [32].

CZA has demonstrated a favourable tolerability profile, with no severe AEs reported in the studies reviewed herein. Only mild AEs were observed. Furthermore, no significant interactions with immunosuppressive therapy were reported.

Real-world clinical experience with CZA in LUTR, particularly in CRE infections, remains limited. While available data support the efficacy and safety of CZA in treating P. aeruginosa infections, evidence specifically about LUTR remains scarce. Further studies are warranted to evaluate the use of CZA in this population. Additionally, close monitoring is advised during treatment, especially for the potential emergence of CZA resistance in K. pneumoniae producing KPC-2 and KPC-33 [33, 34].

Meropenem/Vaborbactam (MVB) combines meropenem with vaborbactam, a novel non-β-lactam BLI.

The primary function of this agent is to treat Enterobacterales that produce KPC enzymes, including those harbouring KPC genes that confer resistance to CZA [35, 36]. In a comparative analysis involving clinical isolates of KPC-positive Enterobacterales, MVB showed more potent in vitro activity compared to other drugs alone [37]. Moreover, MVB demonstrated the highest susceptibility rates against the majority of MDRGNB in a surveillance study that included patients with HBAP [38].

However, vaborbactam does not inhibit Ambler classes B or D carbapenemases. MVB’s activity against other DTR-Gram-negative varies and its activity against P. aeruginosa, Acinetobacter spp., is generally comparable to that of MEM alone [39].

For patients with normal renal function, the recommended dosage is 2/2 g administered in 3 h every 8 h.

Two phase 3 clinical trials have evaluated the efficacy and safety of MVB: the TANGO I trial and the TANGO II trial [40, 41].

The latter is an RCT evaluating the efficacy and safety of MVB versus BAT in adults with CRE infections. Bacteraemia was the more relevant infection (46.8%) while HABAP/VABP was found in 10.6% of patients. ICH, including two SOT recipients, represented 32% of the total cohort. Considering the characteristics of the infections in the population with microbiological confirmation, the trial showed similar mortality rate after 28 days for patients treated with MVB for HABAP/VABP and bacteraemia compared to with BAT (22% vs. 44%, p = NS).

In another retrospective multicentre study, describing clinical characteristics and outcomes of 126 patients treated with MVB for MDRGNB infections, the most common infections were LRTI (38.1%), and the most common isolated pathogens were CRE (78.6%). Thirty-day mortality occurred in 18.3% of patients (n = 23), but only half of these patients received an appropriate dose of medication based on their renal function. Outcomes were similar between patients with CRE and Pseudomonas spp. isolates [42].

Lastly, in a retrospective study comparing the efficacy of MVB (n = 26) with CZA (n = 105) in patients with CRE infections, the clinical success rate was similar in both groups, with around half of the patients treated with MVB having an LRTI (n = 12) [41].

In both the TANGO I and TANGO II trials, patients receiving MVB experienced fewer side effects than those receiving other treatments. In the TANGO II trial, AEs associated with MVB included diarrhoea, anaemia and hypokalaemia. Interestingly, patients receiving MVB treatment experienced a lower incidence of renal failure than those receiving BAT [40, 41].

In a study comparing the efficacy of MVB and CZA, rates of AEs were similar between the CZA group and the MVB group (34.3% versus 23.1%, respectively; p = 0.27). Nephrotoxicity was the most frequent AE, with rates of 29.2% and 14.3% in the CZA and MVB groups, respectively (p = 0.16).

Considering its broad spectrum of activity and good lung penetration, MVB could be a promising option for LUTR with infections caused by KPC-producing CRE. The use of an adequate dose adjusted to renal function will be a future challenge in using this molecule, to ensure correct drug exposure and minimise AEs and the development of resistance.

Imipenem/Relebactam (I-R) combines imipenem with relebactam, a novel BLI without direct antimicrobial activity, but providing reliable inhibition of many Ambler class A and class C [43] β-lactamases, as well as Pseudomonas-derived cephalosporinase [43]. Relebactam is not active against MBLs or class D oxacillinases [44].

The activity of I-R is similar to CZA against CRPA. In addition, in a small percentage of cases, I-R also showed activity against those bacteria that had developed resistance to C/T and CZA [45]. It remains ineffective against A. baumannii and Stenotrophomonas maltophilia and exhibits limited activity against OXA-48-like enzymes [46]. Some data have highlighted the emergence of I-R resistance during this treatment in patients with P. aeruginosa HABAP/HAVAP previously exposed to other cephalosporins. This mechanism could be due to increased expression or structural changes in the MexAB-OprM and MexEF-OprN efflux pumps [47, 48].

I-R against KPC-producing Enterobacterales demonstrated good in vitro activity. Different in vitro strain analyses reported a 98% susceptibility rate for K. pneumoniae producing KPC. However, a pooled estimation of around 280,000 isolates revealed an I-R resistance prevalence rate of approximately 14.6% (95% CI, 0.116%–0.182%), with rates exceeding 50% observed in many countries worldwide [49].

For patients with normal renal function, the recommended dosage is 1.25 g administered in 30′ every 6 h.

Clinical data on I-R use among LuTx patients is lacking. Safety and efficacy of I-R among patients with VAP or HAP are studied in two Phase 3 non-inferiority trials (RESTORE-IMI-1 and RESTORE-IMI-2).

In the first trial, I-R (n = 21) was compared to colistin plus IMP treatment (n = 10) in patients with IMP-susceptible bacterial infections. Patients were treated for HABAP/HAVAP (n = 11, 35%), cUTI (n = 16, 52%) or cIAI (n = 4, 13%). Favourable overall responses were achieved in both arms (I-R 71%; colistin + IMP, 70%) [50].

The second trial evaluated I-R (n = 246) versus TZP (n = 267) in patients with HABAP/HAVAP. In this trial, critically ill patients were studied, but ICH were excluded. The most common pathogens were K. pneumoniae (25.6%), P. aeruginosa (18.9%), Acinetobacter calcoaceticus-baumannii complex (15.7%), and Escherichia coli (15.5%). I-R was non-inferior to TZP, considering 28-day all-cause mortality (15.9% and 21.3% respectively). Instead, on day 28, all-cause mortality in microbiologically modified intent-to-treat patients with a primary diagnosis of HABAP/HAVAP was lower in the I-R arm than the TZP arm (18.6% vs. 30.8%), and the incidence of relapse/clinical failure was comparable between I-R and TZP (14% vs. 12%). Patients with P. aeruginosa infections had a lower clinical response and a higher 28-day mortality rate in the I-R arm. Nevertheless, both treatment arms had comparable microbiological eradication rates at the end of treatment (67% for I-R versus 72% for TZP). This result requires further interpretation as it may be due to differences between the treatment groups that are unrelated to the causative pathogen, given the limited sample size in the I-R group [51].

Another randomised non-inferiority trial compared I-R and TZP for the treatment of HABAP/HAVAP. Again, ICH were excluded from this cohort. The study confirmed that I-R was non-inferior to TZP in terms of 28-day all-cause mortality (11% vs. 5.9%; non-inferiority p = 0.024). It should be noted that mortality was numerically higher in the I-R treatment group, despite non-inferiority being reached [52].

I-R treatment is generally well tolerated. The most common AEs in registration studies were anaemia (10%), nausea and diarrhoea (8%), and elevated liver enzymes (12%) [50, 51].

Regarding renal toxicity, I-R was associated with a more favourable renal safety profile than colistin-based therapy in RESTORE IMI-1. So far, some drug-drug interactions have been described [53]. Carbapenems (imipenem as well as meropenem) have been linked with an increased risk of seizures, especially with the concomitant administration of certain antiepileptic drugs (e.g., valproic acid), due to a marked decrease in those drugs’ levels. Additionally, concomitant use of ganciclovir requires monitoring due to an increased risk of central nervous system toxicity [54]. These interactions are not associated with the new beta-lactamase inhibitor but are inherent to carbapenems.

The role of I-R among ICH, particularly LUTR, requires further investigation. However, considering the data on patients with pneumonia and its anti-pseudomonal spectrum, I-R could play a promising role in this setting. Caution must be exercised regarding DDI and the potential for resistance to emerge.

Cefiderocol (FDC) is a novel catechol-substituted siderophore cephalosporin. FDC can bind extracellular iron and use iron-regulated outer membrane proteins to gain access to bacteria. FDC can overcome resistance mechanisms due to efflux pumps, ubiquitous in MDRGNB such as P. aeruginosa [55]. Moreover, FDC’s potent activity against MDRGNB is also related to its high stability against various ESBLs and carbapenemases (IMP-1, VIM-2, NDM-1, KPC-2/3, L1, OXA-23) [56]. Moreover, FDC demonstrated in vitro activity against AmpC-overproducing strains, a low affinity for chromosomal AmpC β-lactamases, and a low propensity for temporal induction of AmpC β-lactamases [57].

This translates into a potentially vast activity spectrum against MDRGNB, with data from the SIDERO-WT study showing susceptibility rates to FDC of 99.8%, 99.9% and 96% for clinical isolates of Enterobacterales, P. aeruginosa and A. baumannii, respectively [58]. A recent systematic review reported slightly less favourable proportions, with susceptibility rates to FDC of 97%, 91.2%, and 96% for Enterobacterales, P. aeruginosa, and A. baumannii, respectively. Of note, FDC-resistance was significant in NDM-producing Enterobacterales (38.8%, 95% CI 22.6%–58.0%), NDM-producing A. baumannii (44.7%, 95% CI 34.5%–55.4%), and CZA-resistant Enterobacterales (36.6%, 95% CI 22.7%–53.1%), suggesting a cautious use against these microorganisms [59].

The recommended dosage is 2 g administered in 3 h every 8 h for patients with normal renal function.

A growing bulk of evidence is accumulating regarding the use of FDC in IC hosts, including LUTR (Table 4).

Persaud et al. presented a single-centre, retrospective description of FDC among 15 LUTR. FDC was initiated at a median of 105 days post-transplant, with treatment courses ranging from 1 to 93 days. MDR P. aeruginosa was the target pathogen in 13 cases. Of the 15 patients, 14 underwent FDC susceptibility testing, with three yielding an intermediate result, despite no prior exposure to the agent. Overall, 30-day mortality was 26% [60].

In the CEFI-ID study, an analysis of FDC use in 114 ICH adults treated for MDRGNB infections, LUTR constituted 15% (17/114) of the study population. LRTI were the most common infection (55/114, 48.2%), and P. aeruginosa (51/114, 56%) was the most common pathogen. At day 28, clinical success was achieved in 53.3% of cases, and overall mortality was 37.7%.

The PERSEUS study, an analysis of the 261 Spanish patients with severe infections due to MDRGNB (excluding Acinetobacter spp.) enrolled in the FDC early access program, included a relevant proportion of SOT recipients (34/261, 13%). The most represented pathogen was P. aeruginosa (174/261, 66.7%), with 99 (76.7%) isolates resistant to both CZA and C/T. Interestingly, the highest 28-day mortality (27.2%) and the second lowest clinical cure rate (76%) were reported among patients with LRTI, the most common infection site (47.9%), suggesting how infections of this compartment are particularly relevant in shaping the clinical course of patients [61].

Finally, in a post-hoc analysis focused on ICH of the first collected data from the prospective, multicenter national CEFI-SITA study, 84 ICH cases were compared to non-ICH cases. Thirty-day mortality was comparable between ICH and non-ICH (40.8%, 95%CI 27.9%–56.8% vs. 33.3%, 95% CI 22.9–46.9; p = 0.5430). In the multivariable analysis, ICH status and its groups were not associated with higher mortality [63].

FDC is primarily eliminated unchanged in the urine and is not extensively metabolised by the liver [64]. FDC may cause renal impairment [65]. Therefore, it is necessary to closely monitor renal function [64]. Finally, therapeutic and supratherapeutic doses of FDC had no apparent clinically significant effect on the QTc [66].

In the real-life experiences mentioned above, AEs related to FDC were rare or not reported. In the PERSEUS study, 7/314 patients (2.2%) experienced a suspected drug-related AE during FDC administration, and three patients discontinued the drug [60]. In the CEFI-SITA study, 4/200 patients (2.0%) experienced a suspected AE during FDC administration, and FDC was discontinued in two patients [63].

FDC is a solid addition to the therapeutic armamentarium against MDRGNB in LUTR, and scientific societies have endorsed its use for some conditions. In the field of LuTx, FDC is likely a valid solution for treating LRTI due to MDR P. aeruginosa and an alternative for infections caused by MBL-producing Enterobacterales. Further studies are required to understand the need for combination therapy when used in difficult-to-reach sites, such as the lung or the abdomen, and the emergence of resistance.