Due to the rapid variability of Streptococcus pneumoniae strains and serotypes, it is difficult to develop a “universal” phage that would lyse most strains. Moreover, there is a limited amount of data available in the literature that well characterizes phages against S. pneumoniae. Therefore, research on phage therapy primarily focuses on phage endolysins – enzymes that degrade the bacterial cell wall. Unlike antibiotics, lysin can also destroy the cell walls of non-growing bacteria. Additionally, lysin is highly specific to bacterial species, as well as to the phages that produce it (Starkevič et al., 2015). So far, several phage lysins acting against S. pneumoniae strains have been investigated (Cpl-1, Cpl-7/Cpl-7S, Cpl-711, ClyJ, SP-CHAP, and 23TH_48) (Loeffler et al., 2001; Grandgirard et al., 2008; Doehn et al., 2013; Díez-Martínez et al., 2013, 2015; Corsini et al., 2018; Yang et al., 2019; Van Der Kamp et al., 2020; Silva et al., 2021; Alreja et al., 2024; Vander Elst et al., 2025).

Lysin Cpl-1 is an enzyme with hydrolase activity produced by some of the S. pneumoniae phages, which is capable of degrading peptidoglycan – the main component of the bacterial cell wall (Grandgirard et al., 2008). In the studies by Grandgirard et al (Grandgirard et al., 2008), recombinant Cpl-1, specific for S. pneumoniae, was evaluated for antimicrobial treatment in experimental pneumococcal meningitis using Wistar rats. Lysin Cpl-1 showed the ability to kill all of the tested serotypes of S. pneumoniae in vitro. It was also demonstrated that Cpl-1 has an ability to kill S. pneumoniae in an infant rat model of bacterial meningitis when injected intracisternally or intraperitoneally. An intracisternal injection of lysin Cpl-1 decreased the CSF bacterial counts to below the detection limit 30 min after the injection. This effect lasted up to 3 h, and after that bacteria were detected again. An intraperitoneal injection resulted in a constant Cpl-1 level in plasma and CSF for a longer period of time after the injection, which might suggest that this route of administration is more adequate. In summary, it was concluded that the Cpl-1 lysin may be promising as an alternative treatment option for pneumococcal meningitis (Grandgirard et al., 2008). Doehn et al (Doehn et al., 2013). tested the effect of a single dose of recombinant, aerosolized Cpl-1 (0.4 mg) on mice with severe pneumonia caused by pneumococci and observed its effect through 10 days. Endolysin effectively reduced the number of bacteria in the lungs and prevented bacteremia. Although the levels of inflammatory cytokines increased shortly after Cpl-1 inhalation, the mice quickly recovered, as evidenced by weight gain, and the inflammatory infiltrates in the lungs subsided, leading to an 80% reduction in mortality. Díez-Martínez et al (Díez-Martínez et al., 2015). obtained new chimeric phage endolysins (Cp1-711), with enhanced bactericidal activity, by exchanging structural components of two pneumococcal phage enzymes: Cpl-1 (the best lysin tested so far) and Cpl-7S. The bactericidal activity of the new chimeric lysins was tested on bacteria by adding purified enzymes at various concentrations to bacterial suspensions, and living cells were measured after 1 hour. The bactericidal ability of Cpl-711 was tested in a mouse bacteremia model after the survival of mice following injection of various amounts of the enzyme (25–500 μg). The ability of Cpl-711 to limit pneumococcal biofilm formation was also examined. The chimera Cpl-711 significantly enhanced the killing activity of the parent phage lysins, Cpl-1 and Cpl-7S, against pneumococcal bacteria, including multi-resistant strains. Specifically, Cpl-711 at a concentration of 5 μg/ml killed ≥7.5 log of the R6 pneumococcal strain. Cpl-711 also reduced pneumococcal biofilm formation and killed 4 logs of the bacterial population at a concentration of 1 μg/ml. Mice infected intraperitoneally with the pneumococcal strain D39_IU were protected after a single intraperitoneal injection of Cpl-711 one hour later, providing about 50% greater protection than with Cpl-1, which action resulted in a lower number of mice rescued from the lethal infection. This enzyme is a perfect example of a promising therapeutic prospect for the treatment of multi-drug-resistant pneumococcal infections, and swapping domains between phage lysins allows the construction of new chimeric enzymes with high bactericidal activity and different substrate ranges (Díez-Martínez et al., 2013). Alreja et al (Alreja et al., 2024). presented an endolysin targeted against Streptococcus pneumoniae, which also shows greater activity than Cpl-1, the best-characterized pneumococcal endolysin. Cysteine and histidine-dependent amidohydrolase/peptidase (CHAP) domains are widely represented in bacteriophage endolysins but have never been described in pneumococcal endolysins. SP-CHAP exhibited antimicrobial activity against all tested S. pneumoniae serotypes, including encapsulated and non-encapsulated pneumococci, and proved to be more active against 3 of 5 strains than the most studied pneumococcal endolysin. A colony-forming unit (CFU) assay showed a dose-dependent lytic activity, with a 7 log₁₀ decrease in bacterial load after treatment with the highest concentration (100 µg/mL). Surprisingly, the Cpl-1 was able only to cause a 2 log drop. Its efficacy in combating pneumococcal biofilms in vitro and in a mouse nasopharynx colonization model was also demonstrated. In both models, the enzyme caused a greater reduction of bacterial count than when treated with Cpl-1. These findings highlight the therapeutic potential of SP-CHAP as an alternative to antibiotics in the treatment of S. pneumoniae infections. Van der Kamp et al (Van Der Kamp et al., 2020). identified two new Streptococcal phages from the oral microbiome, 23TH and SA01. Their lysins, 23TH_48 and SA01_53, were recombinantly produced, characterized, and tested for their activity against S. pneumoniae strains. They demonstrated that 23TH_48 has broader lytic activity beyond the S. infantis strain, with several S. pneumoniae isolates being highly sensitive to its lytic activity. Given this activity, 23TH_48 may prove to be a promising candidate for helping to combat pneumococcal infections. However, none of these new enzymes were tested in vitro or in clinical conditions. The possibilities of treating S. pneumoniae infections were also studied in so-called combination therapy, in which synergy between antibiotics and phage lysins was observed. For example, earlier mentioned lysin Cpl-1 is a promising addition to the standard antibacterial therapy because of its synergistic activity with penicillin, gentamicin, and Pal/LytA (other phage lytic enzymes) (Rodriguez-Cerrato et al., 2007; Grandgirard et al., 2008; Bustamante et al., 2010; Corsini et al., 2018). Systemic administration of Cpl-1 in combination with penicillin in mice infected with penicillin-resistant pneumococci effectively reduced bacteremia, and thanks to efficient penetration through the blood-brain barrier, it eliminated the bacterial load in the brain, resulting in an increase in survival (89%) with an asymptomatic course of infection. These findings strongly suggest that Cpl-1 may enhance antibiotic sensitivity in β-lactam- and macrolide-resistant S. pneumoniae, serving as a valuable complement to standard antibiotic therapy in cases of multidrug-resistant invasive pneumococcal disease (Vander Elst et al., 2025). Research has also been conducted on improving the stability and transport of lysins to S. pneumoniae. Silva et al (Silva et al., 2021). encapsulated the MSlys endolysin in deformable liposomes for targeted lysin delivery in the treatment of pneumococcal otitis media. Preclinical studies have shown that phage derivatives can effectively reduce S. pneumoniae colonies in in vitro and in vivo models and act synergistically with certain antibiotics (Djurkovic et al., 2005; Vouillamoz et al., 2013; Vander Elst et al., 2025).

Gatea et al (Cheung et al., 2021). conducted a study in which they tested phage preparations against several bacteria, including H. influenzae, in a mouse HIF-1 bacterial sepsis model. Researchers demonstrated four bacteriophages extracted from sewage water with different activity against H. influenzae strains from their environment and other isolates (Gatea Kaabi and Musafer, 2020). Additionally, mixing the examined four phages increased the effectiveness of phage therapy to 100%, which is why the phage cocktail was then successfully used in cases of neonatal sepsis.

Adamczyk-Popławska et al (Adamczyk-Popławska et al., 2020). described a lytic gene set (holin endolysin) of phage HP1, which may be useful as a component of phage therapy or as an enzymatic component destroying H. influenzae bacteria. The activity of this phage may be attributed to the presence of lys and hol genes, products of which are accountable for the signal-arrest-release (SAR) mechanism of endolysin and pinholin, respectively (Starkevič et al., 2015). Researchers also discovered that pinholin has regulatory power over the activity of SAR-endolysin and thus plays a crucial role in the mechanism of cell lysis. That discovery supports the claim that for the best therapeutic outcome, a combination of both of those enzymes should be used. However, all tests were performed in vitro, implying the need for enzyme examinations in more models.

Interestingly, there are vaccines based on phages and their phagemids. In a brought-up case, phagemid [pBSKS::Φ6fm(Hin)] particles incorporated in H. influenzae were used as a mediator for a vaccination, later used to immunize rabbits. H. influenzae Rd30 cells release phagemids that contain their proteins, leading to their appearance and production of IgG antibodies that recognize both the host’s and the phages’ proteins. As it is said, phage NgoΦ6 can form phagemid particles from any Gram-negative bacteria, leading to immunization for that bacteria (Piekarowicz et al., 2022).

Bacteriophages against Klebsiella pneumoniae strains are an area of rapidly developing research, particularly in the context of the increasing antibiotic resistance of this pathogen. The effectiveness of phage therapy against diseases such as wounds and soft tissue infections, bacteremia, and pneumonia was studied in mouse models that accurately reflect the range of diseases caused by these bacteria in humans (Vinodkumar et al., 2005; Kumari et al., 2010; Cao et al., 2022). Kumari and colleagues evaluated the therapeutic potential of five K. pneumoniae phages isolated from sewage material (Kumari et al., 2010). One of the phages, Kpn5, showed the highest lytic activity in vitro and was tested in burn wound BALB/c mouse models (Kumari, 2009; Kumari et al., 2010). It was found that a single intravenous dose of phages at MOI = 1, immediately after bacteria application, was able to save 96.66% of infected mice. However, the delays (6, 12, 18, and 24 h) in the administration of phage resulted in the lower survival rate, with only the 24 h delay resulting in no survivals. Despite that, the observed protection effect in other delays stabilized on day 4, and there were no changes in mortality to the end of the experiment (day 20).

The combination of antibacterial drugs with phages and phage cocktails is a novel approach to reducing bacterial resistance. Many researchers use this new approach in treating phage-resistant K. pneumoniae strains (Verma et al., 2009; Gu et al., 2012; Chadha et al., 2016). Chadha et al (Chadha et al., 2016). observed greater efficacy of the phage cocktail, containing phages Kpn1, Kpn2, Kpn3, Kpn4, and Kpn5, than single phages in treating burn wound infection in BALB/c mice. The reduction of bacterial load occurred in a shorter time compared to single phages, and mice from the cocktail-treated group had the highest wound contraction (~ 91.66%) at the end of the experiment (day 20). Verma and colleagues (Verma et al., 2009) applied a combination of ciprofloxacin and lytic phages in the treatment of K. pneumoniae biofilms. The authors found that this combination reduces the development of K. pneumoniae strains resistant to ciprofloxacin and phage-resistant strains and has greater anti-biofilm activity than monotherapy. A phage cocktail (a combination of three lytic phages GH-K1, GH-K2, and GH-K3), which have different but overlapping host specificities, was used in the treatment of bacteremia caused by the K. pneumoniae K7 strain in a mouse model. The results obtained for phages injected intraperitoneally 30 min after the K7 challenge suggest that the phage cocktail reduced the production of phage-resistant K7 variants. Moreover, the phage cocktail administered to mice with bacteremia increased mouse health and survival rates compared to the control or phages alone (Gu et al., 2012). In the cocktail-treated group, only 1 of 5 mice developed a severe infection, while for the phages alone, 14 of 15 mice developed a severe or moribund state. Due to the possible hematogenous seeding to CSF, such experiments are the second-best studies (after direct meningitis studies) that can be used to estimate the phages’ viability in treating K. pneumoniae meningitis. Thus, often representing the promising results in bacteria eradication and saving the lives of even 100% of infected mice (Table 4).

The use of peptides derived from phages instead of the whole phage represents another alternative therapy option for multidrug-resistant K. pneumoniae. A polysaccharide depolymerase derived from KP36, named depoKP36, was identified, cloned, and expressed in E. coli (Majkowska-Skrobek et al., 2016). This enzyme produced a halo in a spot test on a bacterial biofilm on an agar plate. Furthermore, the efficacy of this enzyme was evaluated in the treatment of infections in G. mellonella. All larvae died without treatment, whereas following post-infection enzyme treatment, up to 40% of the larvae survived.

Although currently there are no studies using the aforementioned phages to treat specifically meningitis, there are various cases where the combination of antibiotics and phages/phage cocktails was used to treat K. pneumoniae. Most of them bring forward rather promising results—the combination of these two types of antimicrobials leads to a total eradication of K. pneumoniae or resolution of symptoms. A Dutch clinician successfully treated a kidney transplant patient with recurrent urinary tract infection caused by ESBL-producing K. pneumoniae bacteria using a combination of phages and meropenem (González-Menéndez et al., 2018). Treatment with meropenem alone failed. The patient received phage therapy from the Eliava Institute (Georgia) and remained infection-free after the therapy. This outcome demonstrated that the combination therapy was effective (Kuipers et al., 2019). Italian clinicians also effectively used a phage cocktail in the treatment of invasive MDR K. pneumoniae infections containing KPC-3 (ST307). At the same time, phages were applied for the decolonization of bacteria from the intestines, which is a key step in preventing future infections and the development of resistant bacteria. The results showed that this treatment has no adverse effects (Corbellino et al., 2020).

Although bacteriophages are not currently used in the treatment of human L. monocytogenes infections due to the intracellular nature of the bacilli, they are used as an alternative or supplement to conventional disinfection methods, both in healthcare and the food industry, to prevent foodborne listeriosis (Palma and Qi, 2024). Products at risk include, among other things, pasteurized whole milk, cheeses, ice creams, fruits, vegetables, and various (both raw and ready-to-eat) meat and fish products (Schlech, 2019). Phage-based solutions, such as Listex™ and ListShield™, have already acquired GRAS (Generally Recognized as Safe) status and are widely used in the USA.

Studies on the preparation ListShield™, a commercially available bacteriophage cocktail that specifically targets Listeria monocytogenes, showed a reduction of L. monocytogenes presence in lettuce, cheese, apples, and smoked salmon, as well as in frozen ready-to-eat meals. Additionally, the ListShield™ preparation removed L. monocytogenes from naturally contaminated smoked salmon while not altering the organoleptic properties of the ready-to-eat meats (Perera et al., 2015). ListShield™ showed promising results, with a study done on Spanish dry-cured ham suggesting reduction below the detection limit (10 CFU/cm²) for lower contamination levels (10³ CFU/cm² and 10⁴ CFU/cm²) and reduction by 3.5 log units for the higher contamination level (10⁵ CFU/cm²), both after storage at 4 °C for 14 days (Grigore-Gurgu et al., 2024).

The preparation Listex, on the other hand, contains the bacteriophage P100, which was originally isolated from wastewater from a dairy plant and selected based on its ability to lyse L. monocytogenes. It belongs to the family Myoviridae, subfamily Spounavirinae, and genus Twortlikevirus and has a DNA genome of 131 kbp with 174 open reading frames (ORFs) encoding proteins (GenBank reference: DQ004855) (Carlton et al., 2005; Klumpp and Loessner, 2013). Phage P100 was developed as a product that obtained GRAS status granted by the FDA/USDA for use in all food materials. This phage was also used in numerous studies demonstrating its effectiveness in removing Listeria contaminants from fish, eliminating Listeria biofilms from stainless steel surfaces, and preventing the presence of Listeria in cooked ham (Holck and Berg, 2009; Soni and Nannapaneni, 2010b, 2010a; Grigore-Gurgu et al., 2024). In 2016 the EFSA Panel on Biological Hazard released an evaluation of the safety and efficacy of Listex™ (EFSA Panel on Biological Hazards (BIOHAZ), 2016). EFSA concluded that bacteriophage P100 doesn’t pose a toxicological risk, as it’s strictly lytic to bacteria and unable to transduce bacterial DNA. Moreover, although not many studies were conducted on naturally contaminated food (only 1 of the 33 considered during the evaluation), Listex™ proved mostly effective on the artificially inoculated ready-to-eat (RTE) samples (in 10^9 PFU/cm² concentration). It’s worth noting that EFSA clearly advises for Listex™ to be used as an additional tool to GHP and GMP, not as a standalone solution, as the product does not fully eliminate L. monocytogenes from samples with concentration levels of 100-10,000 CFU/g. Even though the conclusions of the evaluation seem promising, EFSA stated that more studies on naturally contaminated RTE foods should be undertaken before the product is authorized and used. Additionally, it is underlined that if the product were to be authorized and used, FBO should have additional means to control its dose and L. monocytogenes susceptibility to P100.

Biocontrol of L. monocytogenes using lytic bacteriophage preparations, such as ListShield™ or Listex, may offer an environmentally friendly, green approach to reducing the risk of listeriosis associated with the consumption of various food products that may be contaminated with L. monocytogenes. Although there were attempts to bring Listex™ to the EU market, the efforts have reached a standstill due to current regulations. Despite EFSA’s 2016 review, which shed a cautiously positive light on Listex™’s future in the European Union, the Standing Committee on Plants, Animals, Food and Feed did not pursue the approval process for Listex™, neither in the “decontaminant” nor in the “non-decontaminating processing aid” category (EFSA Panel on Biological Hazards (BIOHAZ), 2016). The convoluted legal route for now ended with a preliminary ruling (22 February 2024) that stated that Listex™ P100 still requires the European Commission’s approval for it to be distributed and used in the member states.

Phages can also be used to disinfect surfaces and equipment in food processing facilities in situations where conventional agents are ineffective or toxic. Their selectivity means they do not destroy biofilms indiscriminately—only those formed by Listeria. In a study conducted by Reinhard et al., various surfaces (stainless steel, thermoplastic conveyor belts, and epoxy floors) were tested for L. innocua reduction using phage P100 under different conditions: temperature (4 °C and 20 °C), phage concentration (1% and 5%), and exposure time (1 hour and 3 hours) (Reinhard et al., 2020). Listeria reduction was observed on all three materials under all tested conditions.

Most research on phage therapy for staphylococcal infections is based on animal models or in vitro studies, while the number of well-documented clinical studies in humans is still limited. Bacteriophages against S. aureus can be found in nature, for example, in nasal swabs, soil, sheep feces, and even in cows with mastitis (Teng et al., 2022). They can be administered individually or in cocktails, such as the StaphLyse™ comprising SAML-4, SAML-12, SAML-150, SAML-229, and SATA-8505 phages (Brouillette et al., 2023). The difference between the efficacy of individual phages and phage cocktails is clearly visible among many studies. For example, one study showed that single phage samples (APTC-SA-2, APTC-SA-4, APTC-SA-12, or APTC-SA-13) are effective against 80%-95% of examined S. aureus isolates, while a phage cocktail (APTC-C-SA01, composition as above) was active against over 98% (Liu et al., 2022). Aside from that, the conjunction of phage therapy and antibiotic therapy shows their synergism - which, in the case of, e.g., phage Sb-1 and oxacillin, increases its effectiveness to 90% from 35% of Sb-1 phage alone (Simon et al., 2021). This synergistic effect was first discovered when sublethal portions of antibiotics led to higher production of lytic bacteriophages. In addition, phages lower bacteria’s MIC, increasing their susceptibility to antibiotics. Moreover, in the mouse model using S. aureus phages, TNF-α and interleukin-6 levels were more notably reduced than in the ceftiofur sodium therapy model, which points to higher effectiveness of phage therapy in this anti-inflammatory aspect (Liu et al., 2024).

Capparelli et al (Capparelli et al., 2007). used phage M^Sa in BALB/c mice infected with S. aureus (including MRSA strains) intravenously. Direct application of phages by the same route resulted in a 100% survival rate when the highest dose was used (10⁹ PFU/mouse). Moreover, the administration of phages 10 days after infection resulted in the clearance of bacteria in the spleens, kidneys, hearts, and blood, while they were still detected in the control group on day 20. In another study, phage therapy was applied in a clinically relevant rabbit model with fracture-related infection caused by S. aureus (Onsea et al., 2021). The intraoperative application of phage in saline was highly effective in preventing infection, as six out of eight animals (75%) were not infected at euthanasia (day 28), while in the control group, only one out of seven animals (14%) was infection-free. However, some difficulties were also noted in treating established staphylococcal infections. The first one is the appearance of phage-neutralizing antibodies, whose count depends on the virus administration route. The second, however, is associated with biofilm and implants, for which treatment phages were loaded in hydrogels. The low effectiveness of this solution might be due to the low titer of phage in the gel (10⁷ PFU/ml), which underlines that the phage particle carriers should allow the high concentration of virions to be gained.

Takemura-Uchiyama et al (Takemura-Uchiyama et al., 2014). investigated phage therapy in a mouse model of sepsis originating from the lungs using phage S13’. Intramuscular administration of the phage 6 hours after infection reduced the severity of the infection and saved 67% of infected mice on day 5, while only 10% of the control group survived. Importantly, the ratio did not change to the end of the experiment. Similarly, Prazak et al (Prazak et al., 2019). demonstrated that intravenous administration of a phage cocktail was as effective as teicoplanin (58 and 50% survival, respectively) in improving survival and reducing bacterial counts in the lungs of rats infected with a methicillin-resistant strain (MRSA). Although, combining antibiotics with phage therapy did not further improve outcomes. Another research showed that nebulization of phages can also reduce the number of bacteria in the lungs of MRSA pneumonia rat model (Prazak et al., 2022).

Bacteriophages may also be an effective local therapy against wounds infected with S. aureus biofilm. Researchers showed that in biofilm-infected wounds caused by S. aureus strains, after applying sharp debridement and bacteriophages, all measured wound healing parameters significantly improved and bacterial counts decreased (Seth et al., 2013). Additionally, some studies show that the application of bacteriophages in S. aureus infections might yield better results when combined with supplements like lactoferrin and that they do not influence the viability of skin cells (Kosznik-Kwaśnicka et al., 2023; Grzenkowicz et al., 2025; Piechowicz et al., 2025). Studies have also shown that phage therapy can be useful in the management of biofilm created on surfaces, which is important due to an increased risk of blood contamination with these bacteria during various medical procedures, such as, e.g., catheterization. Moreover, antibiotics administered on the surface after phage application are often more effective than those administered without phages (Atshan et al., 2023; Liu et al., 2024). Also, as previously established, phages are not only powerful antimicrobial factors themselves but can also be the source of antibacterial bacteriophage-derived enzymes. An example of lysin produced by phages active against S. aureus is Exebacase (Lysin CF-301), which exhibits an anti-biofilm effect that improves when combined with lysostaphin (Schuch et al., 2017).

Due to the ongoing rise of antibiotic-resistant strains (MDR-TB, XDR-TB), phage therapy has once again become a potentially crucial part of TB therapy. Although studies on this topic are still limited, many different approaches to the clinical applications of mycobacteriophages have already been described. Some studies focus on individual phages like D29 or DS6A (Singh et al., 2023; Yang et al., 2024), while other underline the need for a curated phage cocktail, which would help maximize the tuberculocidal effect while minimizing the chance of resistance emergence (Guerrero-Bustamante et al., 2021).

LysB, an endolysin produced by phage D29, effectively cleaves the ester bonds in the mycolic acid layer of the mycobacterial cell envelope, enabling the lysis of M. tuberculosis in vitro (Abouhmad et al., 2020; Singh et al., 2023). Studies show that both the phage itself and the purified LysB might be valuable additions to other phages and to rifampicin (and isoniazid), as the synergistic effect was observed in vitro in both cases (Guerrero-Bustamante et al., 2021; Singh et al., 2023). Although D29 shows promising results, a recent study suggests that its efficacy in agar plates doesn’t positively correlate with its efficacy in broth culture, which might lead to problems later on with using D29 on mice or more humanized models (Yang et al., 2024). This study also demonstrates phage DS6A as the better alternative, because it lysed M. tuberculosis both in agar plates and in liquid culture. Additionally, DS6A was tested on Mtb in primary human macrophages, which ended in a successful bacilli eradication within seven days. That result was further positively translated into a humanized mouse model treated intravenously with DS6A, which showed reduced bacterial loads in infected organs, improved pulmonary function, and weight gain compared to untreated controls. Due to the route of DS6A administration, phage distribution varied, with DS6A gathering in higher amounts in the spleen compared to the lungs, suggesting the need for alternative delivery methods to optimize lung clearance.

Aside from individual phages, researchers also explored the possibility of using various phages together to achieve the best possible clinical outcome. In a study conducted at the University of Pittsburgh, researchers have selected and tested out various phages, both the naturally lytic mycobacteriophages and particularly engineered derivatives of temperate mycobacteriophages, to finally choose the final set for the cocktail (Guerrero-Bustamante et al., 2021). Their selection criteria included broad host range, minimized risk of resistance and cross-resistance, lytic tuberculocidal activity, synergy with antibiotics, and ease of propagation. That way five phages broke forth—Adephagia Δ41Δ43, D29, Fionnbharth Δ45Δ47, Fred313_cpm Δ33, and Muddy_HRMN0157-2. However, the authors admit that this cocktail is likely to be refined before clinical evaluation, which underlines the need for more studies to be done on the final selection of the most adequate phages.

Bacterial meningitis caused by obligatory anaerobes is a rare (<1%) but potentially severe central nervous system (CNS) infection. It requires prompt diagnosis and targeted treatment. The most isolated obligate anaerobes in meningitis are Bacteroides fragilis, Fusobacterium necrophorum, Clostridium perfringens, and Cutibacterium acnes.

Bacteroides fragilis typically requires specific conditions (post-surgery state, trauma, immunodeficiency) to induce bacteremia grave enough to attack the CNS (Zafar and Saier, 2021). That’s why anaerobic meningitis caused by B. fragilis is a rather rare occurrence, especially in adults (Feder, 1987). Nevertheless, with the rise of antibiotic resistance, alternative antimicrobial therapeutics are growing in importance, renewing researchers’ interest in possible phage solutions (Centers for Disease Control and Prevention (CDC), 2013; Merchan et al., 2016; Yekani et al., 2022). For now B. fragilis phages are mentioned in literature mostly as the means of monitoring the fecal contamination of water (Queralt et al., 2003), with only limited studies on possibly therapeutic phages. One study focuses on the phage vB_BfrS_23 (Tariq et al., 2020), but due to a very narrow host range, its therapeutic usefulness might be limited. Nevertheless, a more recent study characterizes GEC_vB_Bfr_UZM3 as a possible therapeutic bacteriophage against B. fragilis. UZM3 is a strictly lytic phage able to maintain stability for around 6 hours at body temperature and in body pH environments. Depending on the strain on which UZM3 was propagated, the phage’s host range varied from five to nine bacterial isolates out of 15, which translates into a possibly higher clinical value of UZM3 compared to vB_BfrS_23 (Bakuradze et al., 2023). Another therapeutically promising phage is vB_BfrS_VA7 (VA7), which is characterized by its virulent nature and its ability to effectively reduce B. fragilis cell counts in the cell culture by 2 logs (Bakuradze et al., 2021). Moreover, the same study suggests that the immunomodulating effect of VA7 (lowering the mean level of IL-8, compared to untreated cell culture) might be additionally advantageous in a potential clinical setting. Unfortunately, due to the scarcity of studies on all of the phages listed above, there is still a long way to go before any of those phages are used to treat B. fragilis anaerobic meningitis.

Meningitis caused by Fusobacterium necrophorum is also rather rare, which translates into the number of studies being published - only two relevant studies touch on the topic of F. necrophorum phages. The first one describes the isolation of phage FnP1, which proved effective on biovar A (5 out of 7 strains) and B (7 out of 7 strains) of F. necrophorum, but not on biovar C (Tamada et al., 1985). Unfortunately, that phage wasn’t studied further, so it’s hard to establish its relevance in a potential therapeutic setting. The second study, a fairly recent one, explores six novel phages - φFN37, φRTG5, φKSUM, φHugo, φPaco, and φBB (Schwarz et al., 2024). Among these phages, two (φKSUM and φBB) showed the most promise for therapeutic application against F. necrophorum, as both infected and suppressed growth of the F. funduliforme subspecies. Additionally, even though all six phages contained genes associated with lysogeny, φKSUM and φBB did not form lysogens, making them stronger candidates for potential biocontrol. To fully explore the potential of F. necrophorum phages as viable alternatives to antibiotics, further investigation is still needed.

Even though Clostridium perfringens doesn’t commonly cause meningitis, many studies explore the possible use of its phages in the clinical context due to other aspects of its pathogenicity. Most C. perfringens phages are explored in the context of poultry treatment, but their use in human infections might still be possible when adequate studies are undertaken. One study summarizes the twenty phages characterized more thoroughly up until 2022, with six of them used successfully to reduce necrotizing enterocolitis-associated mortality in broiler chickens by 92% and to control necrotic enteritis in them (Miller et al., 2010; Bae et al., 2021; Venhorst et al., 2022). Additionally, it was ascertained that phage phiCJ22 is highly stable under pH 5–9 and at the temperature reaching 60 °C, which makes it a viable phage for potential use in more humanized models. In another study, 32 phages were isolated, out of which five had the broadest host ranges and killed over 90% of the examined C. perfringens strains (Thanki et al., 2024). The last two phages worth mentioning are DCp1 and vB_CpeS_BG3P (BG3P). DCp1 was proved to be effective against C. perfringens biofilm, making it possibly feasible for future therapeutic use (Tang et al., 2023). BG3P, on the other hand, remains stable under pH 2–7 and temperatures up to 50 °C. It is also able to completely inhibit bacterial growth compared to untreated cultures, and produces the endolysin LysCP28, which is characterized by high antimicrobial and antibiofilm activity (Huang et al., 2022; Lu et al., 2023). Despite the promising amount of possibly useful phages, more studies need to be conducted to test their feasibility on humanized models.

Meningitis caused by Cutibacterium acnes is rare and usually associated with neurosurgical procedures, the presence of intracranial devices, or a long-standing shunt. In patients without such factors, cases are very sporadic (Prozan et al., 2020). Preclinical studies on phage therapy for Cutibacterium acnes currently focus primarily on dermatological applications. An important early study focused on the phage PA6, which effectively infects and destroys all 32 Cutibacterium acnes isolates (types I and II) used in the study (Farrar et al., 2007). Other researchers confirmed the activity of the phage PAP 1–1 against pathogen cells, with more promising results obtained when the phage was combined with bacteriocin from Lactococcus lactis CJNU 3001 and nisin (Han et al., 2023). Additionally, another study described phage CAP 10–3 isolated from a human acne lesion and proved the activity of its lysin in a dose-dependent manner (Kim et al., 2023). In the same year, a putative N-acetylmuramoyl-L-alanine type-2 amidase (PaAmi1) from bacteriophage PAC1 was characterized, and its peptidoglycan degradation activity was confirmed (Varotsou et al., 2023). The results of this study suggest that PaAmi1 is a new, effective antimicrobial agent and provide proof of concept that bacteriophage genomes are a rich source of antimicrobial proteins that can be further used to design novel or improved endolysins.

Brucella is a zoonotic bacterium that primarily causes brucellosis but can sometimes cause more complicated conditions, such as meningitis (Mousa et al., 1986). It can affect both humans and animals, and its therapy is currently challenging. Phage therapy for Brucella spp. infections (brucellosis) is a very promising area but still relatively early in the research phase. Because of the intracellular development of this pathogen, different approaches using phages are considered. In a Swiss albino mouse model infected with Brucella abortus, it was found that intravenous administration of the phage (10⁶ PFU/mL) significantly reduced the number of bacteria in the spleen by 2 logs in the treatment group compared to the control group (Prajapati et al., 2014). In a study described by Jain et al (Jain et al., 2017), phage lysate of B. abortus was used for immunization of guinea pigs, resulting in satisfactory protection against the pathogen comparable to a vaccine. Moreover, transfer of immunized guinea pigs’ serum to mice also caused sufficient immunization. In another study, Brucella strain 19 was used as a carrier to deliver the phage inside the phagocytes where this pathogen resides during an infection (Mohan and Saxena, 2020). The authors suggest that it’s important to apply the mixture of bacteria with phages before lysis can be observed in vitro. Such action allows bacteria with phages to infect phagocytes and then phages to start lysing the carrier. Next, phage progeny infects and lyses Brucella cells from the original infection. The cattle with brucellosis were treated with 2 mL of B. abortus strain 19 vaccine with brucellaphage through the subcutaneous route. The use of bacteriophages resulted in an increase in the antibody levels compared to the prevaccination level. These results underline that phages, despite their inability to infect human cells directly, can be used to treat intracellular infections.

Leptospirosis is a zoonotic infection that causes more than a million human infections annually. Although leptospirosis affects many organs, the form involving the central nervous system (especially aseptic meningitis) is less frequently recognized. In one series of suspected aseptic meningitis cases, it was observed that Leptospira may be responsible for approximately 5−13% of all aseptic meningitis cases in certain regions (Bandara et al., 2021). The potential of phage therapy for leptospirosis remains underexplored. Only a few phages have been identified: vB_LbiM_E1 (LE1), vB_LbiM_LE3 (LE3), and vB_LbiM_LE4 (LE4) (Girons et al., 1990). Unfortunately, among the tested strains, they were active only against the Patoc serovar and the L. meyeri strain 201601301, indicating their narrow host range. Additionally, various mechanisms of phage resistance exhibited by Leptospira, such as the CRISPR-Cas system, type IV restriction–modification (R–M) system, effector protein PrrC, and the single-gene anti-phage system Borvo, might complicate the use of Leptospira phages even more (Petakh et al., 2024). Currently, phages against Leptospira are mostly used not due to their antibacterial activity but because of their ability to display different proteins on their structure, allowing the investigation of interactions between bacteria and cells or even modulation of antibodies (Lima et al., 2013; Mohd Ali et al., 2021).