As a worldwide public health threat, the World Health Organization (WHO) established a list of antibiotic-resistant bacteria in 2017 (Tacconelli et al., 2018). The aim was to prioritize and stimulate research and development strategies of new active drugs. Among relevant criteria such as patient mortality prevalence, health-care burden, and trend of resistance worldwide, carbapenem-resistant Pseudomonas aeruginosa (CRPA) were considered “critical-priority” bacteria (Tacconelli et al., 2018). Indeed, Carbapenem antibiotics are reserved for the treatment of multi-drug resistant (MDR) bacterial infections, and, when bacteria develop resistance to them, treatment options become extremely limited.

P. aeruginosa is an opportunistic pathogen responsible for both severe acute and chronic infections, and is a significant cause of healthcare-associated infections, particularly in critically ill and immunocompromised patients (Figure 1). Since its first description in wound infections (Gessard, 1882), P. aeruginosa is now a well-known pathogen. Pathogenesis of P. aeruginosa is mediated by an arsenal of virulence factors: motility, adherence to biotic and abiotic surfaces, secreted toxins also called effectors that are released in the environment or injected into host cells or other bacteria (Figure 2). These effectors are able to modulate or disrupt host cells signaling pathways, target extracellular matrix, induce tissue damage, and shape the local microbiome by competition (Figure 2). The ability of P. aeruginosa to form a biofilm is also a key factor that increases drug resistance and escape from host defense and is responsible for colony tolerance to disinfectants on medical devices (Mulcahy et al., 2014; Pang et al., 2019; Jurado-Martín et al., 2021). As all these factors contribute to pathogenicity by complementary actions, P. aeruginosa is characterized by a combinatorial multifactorial virulence. Moreover, multiple mechanisms of antibiotic resistance have been identified, including intrinsic membrane permeability, drug efflux systems, production of antibiotic-inactivating enzymes, and loss of porin function (Pang et al., 2019). Finally, the plasticity of its (i) virulence factor gene expression, (ii) antibiotic resistance, and (iii) metabolism in response to selective pressure is one of the most challenging features of P. aeruginosa allowing the transition from acute to chronic infections. Acute infections are mainly associated with planktonic life style, whereas biofilm plays a major role in persistent infections. This remarkable ability of over-adaptation in a dynamic way allows this pathogen to escape immune system and become MDR or extensively drug resistant (XDR). Once a chronic infection in the patient is established, P. aeruginosa is really difficult to treat.

To meet this major treatment issue, research is active with a variety of approaches ranging from disarming the pathogen with an anti-virulence strategy to eliminating it. In parallel with the growing knowledge of the molecular mechanisms of P. aeruginosa–host interaction, the number of potential therapeutic targets is increasing. However, the path to validate efficacy of a new drug in human is long and uncertain. Two antibacterial tracks of development are followed and consist in “traditional agents” (small molecule directly targeting the bacterium for a bacteriostatic or bactericidal action) and new therapeutics called “non-traditional” (large molecule and/or not acting by directly targeting bacterial components essential for bacterial growth). Characterized by new, non-bactericidal, and generally species-specific modes of action, the paradigm is that non-traditional agents are less likely to generate the dreaded resistance (Tse et al., 2017). Indeed a non-killing agent reduces the selective pressure, and its specificity avoids cross-resistance potentially induced by horizontal genes transfer (Tse et al., 2017). This review focuses on new solutions for P. aeruginosa infections that are in active clinical development, i.e., currently being tested in humans and may be approved for patients in the coming years. A literature search was performed in a context of need to address clearly the question: What is new in the anti–P. aeruginosa clinical development pipeline since the 2017 WHO alert?

With more than 74,000 articles published in PubMed^® in December 2021, knowledge about P. aeruginosa is growing in the fields of its virulence, antibiotic resistance, lifestyle, metabolism, clinical manifestations, clinical or observational studies, health and economic burden, among others. Numerous articles or reviews detail the potential new targets from basic research, the emerging new candidate therapies in early phase, or the pipeline of antibacterial molecules in preclinical and clinical testing and the hope that comes with it (Burrows, 2018; Tümmler, 2019; WHO, 2020; Theuretzbacher et al., 2020; WHO, 2021; Yaeger et al., 2021). It is difficult to find one’s way through all this data, which is often mixed and rarely dedicated to P. aeruginosa. It is important to keep in mind that before proof of concept in a phase II study, no efficacy is proven in patients. Efficacy–safety balance evaluation in P. aeruginosa–infected patients is an unavoidable milestone. To dedicate this review only to anti–P. aeruginosa treatments in clinical development, a literature search was conducted with a specific strategy in the PubMed^® database between January 2017 and December 2021. The following criteria were used: (i) keywords related to development (antibacterial pipeline, antibiotic pipeline, and clinical development) and treatment (drug, antibiotic, and therapy), (ii) selection of phase I to III clinical trials, and (iii) exclusion of research data or preclinical targets (Figure 3). Clinical trial registries and websites of companies or organizations with a special interest in the field were consulted to verify or supplement the data (Figure 3).

A vaccine against P. aeruginosa for at-risk patients [i.e., those older than 65 years and those with cystic fibrosis (CF), bronchiectasis, or chronic obstructive pulmonary disease] could reduce the prevalence of infections, the overall burden disease, and the use of antibiotic treatment. Therefore, although formal data are lacking, the vaccine, by limiting the use of antibiotics and thus reducing selection pressure on pathogens, has a significant indirect impact on the emergence of antibiotic resistance (López-Siles et al., 2021; Micoli et al., 2021). An important challenge for obtaining a good vaccine is the identification of ideal antigens, i.e., antigens accessible or presented to the immune system, adequately immunogenic and highly conserved across all serotypes (Sainz-Mejías et al., 2020), regardless of the stage or location of the infection. Antigen variability, plasticity of virulence factor expression during acute to chronic infection described earlier in the introduction, and bacterial localization (the mucosal immune response is compromised in dehydrated and sticky mucus in lungs of CF patients) are a challenge. In addition, it is also necessary to find a safe but immunogenic vector and antigen formulation (with or without adjuvant), allowing mucosal vaccination, intracellular and extracellular delivery of antigens to achieve specific cellular and humoral immunity, as well as long-term immunity. Finally, the choice of the target pathology is also very important: bacteremia, chronic lung infection, keratitis, urinary tract, or skin infection for the clinical trial.

Despite active research for a P. aeruginosa vaccine during over half a century, no vaccine has yet been approved (Sainz-Mejías et al., 2020; López-Siles et al., 2021; Micoli et al., 2021). Several antigens [lipopolysaccharide (LPS), alginate, flagellum, type 4 pili, outer membrane proteins (OMPs), type 3 secretion systems T3SS, T2SS effectors, autoinducers, and iron-uptake proteins] have been targeted in clinical development, but to date, only three vaccines reached the phase III trials (Rello et al., 2017). The investigational vaccine IC43, a recombinant OMPs (OprF porin/OprI lipoprotein)–based vaccine, appeared to be the last promising based on the favorable safety and immunogenicity profile from phase II study (Adlbrecht et al., 2020).

The vaccination (or active immunotherapy) strategy has various limitations, particularly for immunocompromised patients who may not be able to induce an appropriate immune response. Moreover, the immediate need for protection may not be met, including the time required for the development of an adequate immune response after vaccination, as discussed above (Adlbrecht et al., 2020). The uncommon mucus produced in the CF lung can also be a barrier to the systemic and specific immune response against P. aeruginosa that colonize the lower airways. Nevertheless, antibodies represent an effective and specific line of defense of the immune system. Passive immunotherapy with antibacterial monoclonal antibodies (mAbs) is therefore an alternative therapy against P. aeruginosa that could be interesting in certain clinical situations. Indeed, mAbs target specific surface antigens that are not usually the targets of antibiotics and are thus active against MDR bacteria. Antibacterial strategies are developed by mAbs binding to bacterial surface antigens, inducing opsonophagocytic killing by the host immune system or by binding to a virulence factor, such as toxin and thus neutralizing it (Lakemeyer et al., 2018). To date, no mAbs against P. aeruginosa have been approved.

Among the latest anti–P. aeruginosa mAbs, AR101 and AR105 have been developed as adjuvant strategy to antibiotics. Whereas AR101 targets the LPS serotype O11 surface O-antigen, AR105 targets alginate, a polysaccharide required for biofilm formation (Zurawski and McLendon, 2020; Lu et al., 2011). A phase IIa trial evaluating AR101 was completed in 2009 for hospital acquired pneumonia, showing promising results in clinical cure and survival rates (NCT00851435). A phase II trial testing the efficacy, safety, and pharmacokinetic of AR105 in addition to standard-of-care (SOC) antibiotics for pneumonia cause by P. aeruginosa was started in 2017 (NCT03027609). MEDI3902, a bispecific mAb, is able to recognize two different targets of P. aeruginosa, each containing a distinct epitope: the T3SS needle-tip protein PcrV, involved in host cell cytotoxicity and the surface exopolysaccharide Psl, involved in epithelial attachment and biofilm formation (Figure 2; Sato et al., 2011; Ryder et al., 2007). Designated as a Fast Track drug from Food and Drug Administration (FDA), MEDI3902 is under development for the prevention of healthcare-acquired P. aeruginosa pneumonia in high-risk patients¹ . PsAer-IgY (egg yolk immunoglobulin), an avian polyclonal anti–P. aeruginosa antibody, has entered in a phase III trial to test its efficacy in preventing recurrence of P. aeruginosa infection in patients with CF (Thomsen et al., 2016).

Almost 60 years after their clinical approval, polymyxins remain a class of antibiotic available for many MDR Gram-negative bacteria but used as last-line therapeutic option due to their potential human nephro-and neurotoxicity (Li et al., 2019). Polymyxins are small cyclic cationic lipopeptides, interacting with the anionic lipid A component of LPS, in the outer membrane of Gram-negative bacteria, leading to cytoplasmic membrane disruption and bacterial cytotoxicity (Li et al., 2019). Their clinical use has restarted in recent years with polymyxin B and polymyxin E (colistin). On the basis of structure–activity–toxicity relationships, many efforts have been made to modify the original polymyxins and improve their safety profile. SPR741 is a polymyxin B derivative with less nephrotoxicity. SPR741 has no direct antibacterial activity but potentiates the efficacy of co-administered antibiotics (French et al., 2020), which alone would not have access to their intracellular targets (Corbett et al., 2017). SPR741 completed a phase I clinical trial in 2017 (Eckburg et al., 2019).

Among bacterial targets, peptidoglycan (PG) synthesis remains a privileged target with investigations on molecules of the β-lactams family. One field of development focuses on the modification active β-lactams by addition of an iron-chelating molecule, facilitating transport into bacteria (De Carvalho and Fernandes, 2014). Cefiderocol, a first in class of siderophore-cephalosporins, is able to bind to extracellular free iron, allowing active transport into the periplasmic space of Gram-negative bacteria through siderophore uptake systems. Cefiderocol subsequently binds to penicillin-binding proteins (PBPs), inhibiting bacterial PG cell wall synthesis that leads to cell lysis and death (EMA, 2020a). Although siderophore antibiotics have been investigated for several decades, none of them progressed to clinical development because of poor correlation between in vitro activity and in vivo efficacy or because of toxicity (El-Lababidi and Rizk, 2020). Promising assessment of cefiderocol for the treatment of cUTI in patients at risk of MDR Gram-negative infections started in phase II trial in 2015 (Portsmouth et al., 2018).

LPS that constitutes the OM outer layer at the surface of Gram-negative bacteria represents also an attractive target against pathogens (MacNair et al., 2020). By binding to the LPS transport protein D (LptD), the small cyclic peptide murepavadin causes a specific P. aeruginosa LPS biogenesis alteration (Martin-Loeches et al., 2018). A second phase II study of murepavadin co-administered with SOC in VAP due to P. aeruginosa finished in 2017 (NCT02096328).

Among antibiotics, β-lactam targets specifically the PG biosynthesis through covalent binding/interaction to PBPs, the enzyme involved in late stages of PG synthesis. However, bacteria under selection pressure could produce β-lactamases, enzymes who hydrolyzed the β-Lactam ring. Hydrolyzing β-lactam antibiotics has made many of them ineffective becoming a major resistance issue in Gram-negative bacteria (Tümmler, 2019). Consequently, a synergic combination of β-lactam and an appropriate β-lactamase inhibitor (BLI) restoring β-lactam activity is a frequently used strategy. The well-known combination amoxicillin (β-lactam)/clavulanic acid (BLI) is a good example as it represents one of the most prescribed antibiotic worldwide and is a WHO-designed “core access antibiotic” that should be consistently available (Sharland et al., 2018).

Relebactam is a new BLI with activity against a broad range of β-lactamase enzymes including carbapenemases (Smith et al., 2020). In vitro addition of relebactam to imipenem restored imipenem activity against several imipenem-resistant bacteria, including P. aeruginosa (Smith et al., 2020). Relebactam in association with imipenem-cilastatin demonstrated efficacy in phase II trials dedicated to cUTIs (Sims et al., 2017) and complicated intraabdominal infections (cIAIs) (Lucasti et al., 2016), thus entered in phase III trial in HAP/VAP (Titov et al., 2020) versus piperacillin-tazobactam– and imipenem-resistant pathogens versus colistin (Motsch et al., 2020), a polymixin used as last resort therapy. Taniborbactam is a highly potent pan-spectrum new BLI that inhibits all classes of β-lactamase enzymes (Liu et al., 2020). In combination with the fourth-generation cephalosporin cefepime, taniborbactam positively completed the phase I milestone in 2017 and thus supported other trials (Dowell et al., 2020).

Bacteriophages that infect and lyse their target bacteria are being reconsidered as an alternative therapy to treat MDR bacterial infections. Interestingly, some phages are also able to disrupt biofilm barrier with EPS-depolymerase activity (Cornelissen et al., 2012). Phage therapy includes mono-phage, phages-cocktail, phage-derived enzyme (lysin), bio-engineered phage, and phage combined with antibiotics (Patil et al., 2021). Systematic literature search shows a large number of case study reports and compassionate use for severe patients in specialized centers but despite promising results recent clinical trial evidence-based will be necessary. The first randomized controlled trial using a cocktail of natural lytic P. aeruginosa phages for the topical treatment of infected burn wound patients was stopped on January 2017 because of PP1131 insufficient efficacy versus SOC (Jault et al., 2019).

As an essential nutrient for growth and biofilm establishment, P. aeruginosa has developed different strategies to acquire iron from its environment (haem or siderophores mediated uptake systems and porins) (Zhang et al., 2021). Host fights infection reducing iron biodisponibility for its invading bacteria using, for example, transferrin in serum and lactoferrin in mucosal secretion (Zhang et al., 2021). Acting as an iron mimetic, gallium disrupts multiple iron-dependent synthetic and metabolic pathways (Frei, 2020). Preliminary clinical study raised the possibility that gallium may be a safe and effective treatment for human infections (Goss et al., 2018). A phase II study to test IV gallium nitrate to control  P. aeruginosa  infections in adults with CF started in 2016 (NCT02354859). Because the IV form of gallium involves a continuous 5-day infusion, which is a demanding treatment regimen for patients, the AR501 inhaled formulation of gallium citrate self-administered via a nebulizer device was designed to be given once a week, allowing for direct delivery to the lungs¹⁰. Another antimicrobial with an activity on iron metabolism is ALX-009, a combination of lactoferrin and hypothiocyanite, natural major antimicrobial substances deficient in the airway secretions of patients with CF (Tunney et al., 2018). Lactoferrin deprives bacteria of iron due to its iron chelator activity and can bind to cell membranes, leading to alterations in permeability and enhancing bacterial killing by other antibiotics (Rogan et al., 2004). Hypothiocyanite is a highly reactive compound that oxidizes proteins to created disulfide bonds that perturb the bacterial physiology (Thomas et al., 1978). ALX009 started a phase I in 2015 as an inhalable solution administered through nebulization (NCT02598999).

Among virulence factors of P. aeruginosa, biofilms increase bacterial adherence, immune system evasion, and antibiotic tolerance by blocking the diffusion of positively charged drugs. OligoG is an alginate oligosaccharide with the potential to reduce sputum viscosity of patients with CF by chelating calcium (Ermund et al., 2017), easing clearance of mucus from patient airways, reducing microbial burden and inflammation. OligoG was also shown to disrupt biofilm structure of P. aeruginosa mucoid phenotype (Powell et al., 2018) and could in consequence improve host immune system action and the effectiveness of antimicrobial agents. To determine the safety and local tolerability of multiple dose administration of inhaled fragment in healthy volunteers, a phase I study was performed with a particular attention on pulmonary functioning and adverse events (NCT00970346).

As T3SS and associated toxins are major virulence factors of P. aeruginosa (Figure 2), innovative therapeutic strategies are developed to reduce the infection severity. Despite many hopes based on in vitro or preclinical activities, only one treatment targeting T3SS is currently in clinical development (Figure 4E). Ftortiazinon, a small molecule with a strong potential as an antibacterial therapy (Sheremet et al., 2018), developed by the Gamaleya Research Institute, entered a phase 2 in combination with cefepime for the treatment of patients with cUTI caused by P. aeruginosa in 2018 (NCT03638830).

This review aimed at providing an up-to-date picture of therapeutics against P. aeruginosa currently in clinical development, since the 2017 WHO alert. Only one antibacterial drug with a new mode of action has been approved by FDA and EMA against P. aeruginosa (Figure 4B). Among 18 drugs of interest anti–P. aeruginosa in the development pipeline described in this review, only one new combination of β-lactam/β-lactamase inhibitor of antibiotics is in phase III trial (Figure 4C). Derivatives of existing antibiotics considered as “traditional agents” are highly represented (Figure 5). Diverse “non-traditional agents” including bacteriophages, iron mimetic/chelator, and anti-virulence factors are significantly represented but unfortunately still in early clinical stages. There is no vaccine in development to prevent P. aeruginosa infections despite a half century of research effort specifically focused on this challenge (Sainz-Mejías et al., 2020; ClinicalTrials.gov., 2021).

Studying pipeline anti–P. aeruginosa since 2017 up to now shows how development of a new treatment can be a difficult process. Lack of correspondence between in vitro or preclinical study and phase II clinical response questions the choice of pertinent animal models recreating human infection conditions (Theuretzbacher et al., 2019; Theuretzbacher et al., 2020). Methodology used is often a strategic issue, with the crucial definition of the clinically significant primary outcome (Merakou et al., 2018; Adlbrecht et al., 2020) and the infection-site that could best allow to meet efficacy criteria. The absence of statistically significative evidence is not the evidence of absence of efficacy. We can question, for example, if death is a good primary outcome when critically ill patients are entering in ICU (Merakou et al., 2018) and if lung function measure is a pertinent primary outcome for a CF patient with chronical severe disease (Ali et al., 2019; Van Koningsbruggen-Rietschel et al., 2020). Studying pipeline also underlines that a late clinical stage development can be interrupted due to unexpected toxicity¹⁶ or bankrupt of a biotechnology company. COVID-19 pandemic had also an important impact on clinical trials recruitment and timelines during the last 2 years.

P. aeruginosa–host interactions and host immunometabolism are not yet enough understood, complicating the development of effective therapies and vaccines (Sainz-Mejías et al., 2020). Because of the fact that P. aeruginosa is an opportunistic bacterium, patients infected with P. aeruginosa are mostly immunocompromised individuals. This makes it difficult to develop a vaccine for these patients (Yaeger et al., 2021). As P. aeruginosa is characterized by its genomic plasticity, drugs or vaccine must be designed to target conserved elements between strains to ensure an optimal efficacy (Tse et al., 2017). This constitutes a true limitation regarding strains like PA7, a non-respiratory MDR that lacks the T3SS and its effectors or the exotoxin A (Roy et al., 2010) or the Liverpool epidemic strain, a lineage with enhanced virulence and antimicrobial resistance characteristics (Salunkhe et al., 2005).

Each class of treatment against P. aeruginosa presents strengths, weaknesses, opportunities, and threats that have to be taken into account in global clinical development considerations (Table 1). Among “non-traditional agents”, a vaccine could offer a long-term sustainable approach to infection prevention because it will decrease the need for antibiotics and hence the emergence of antibioresistance (Micoli et al., 2021). Anti-virulence strategy may have the potential through novel, specialized, and non-killing modes of action to reduce the selective pressure responsible of MDR (Dickey et al., 2017) or select less virulent strains. Regarding combinatory multifactorial virulence, an anti-virulence factor used alone do not has therapeutic utility and hence must be used as pre-emptive or adjunctive treatment in combination with traditional antibiotics (Dickey et al., 2017). Despite many years of research, no anti-virulent agent has yet been introduced in clinical use against P. aeruginosa. Novel targets for anti-virulence strategy must be proposed and T6SS as a key virulence factor of P. aeruginosa (Figure 2) could be selected. Inhibition of T6SS assembly or neutralization of its toxins (amidases, phospholipases, protease, pore forming toxin, and iron-acquisition effector) (Figure 2) would allow to interfere at two moments of the infectious process: (i) during host colonization (competition for the niche with microbiota or other pathogens in case of CF patients) and (ii) escape from the immune system (internalization and autophagy).

Current antibiotics are less effective due to increasing resistance, and some P. aeruginosa isolates are resistant to all available treatments, underlying the unmet medical need. Given the development duration, the pipeline remains unsatisfactory leading best case to the approval of new antibacterial drugs that treat CRPA in several years. In addition, as the United Nations, WHO, and numerous experts published in an April 29, 2019, report, immediate, coordinated, and ambitious action must be taken to avoid a potentially disastrous antimicrobial resistance crisis. If not, drug-resistant diseases could cause 10 million deaths each year by 2050 warns the UN Interagency Coordination Task Force on Antimicrobial Resistance¹⁶. Developing a new treatment takes years, however, COVID-19 pandemic has demonstrated that it is possible to accelerate the development of a molecule or a vaccine in case of crisis (Krammer, 2020). Beyond the investments needed to build a robust pipeline, the Community needs to reinvent medicine with new strategies of development to avoid the disaster.

SR, AG, and SB wrote the manuscript. All authors contributed to the article and approved the submitted version.

Work of SB team is funded by the Centre National de la Recherche Scientifique, the Aix-Marseille Université, and by grant from the Agence Nationale de la Recherche (ANR-21-CE11-0028-01). AG is supported by “Vaincrela mucoviscidose” (VLM) and “Association Grégory Lemarchal”(grant numbers RF20190502416 and RF20210502854), ANR-15-IDEX-02, SATT Linksium, and Fondation Université Grenoble Alpes. SATT Linksium was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. SR is currently working in the Medical Department of Novartis Gene Therapies France SAS as Senior Medical Science Manager. SR is a PhD student independently of his professional position and on a totally different therapeutic area.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.