The MAPVAP consortium, funded under grant 01KI2124, investigated the mechanistic basis of two lytic phage cocktails targeting multidrug‑resistant *Pseudomonas aeruginosa* and *Escherichia coli* for use in ventilator‑associated pneumonia. The project ran from 1 January 2020 to 30 June 2023 and involved the Charité – Universitätsmedizin Berlin, the Institut Pasteur in France, the Leibniz Institute DSMZ in Germany, the University of Jena, and the Freie Universität Berlin. Charité supplied the phage preparations, performed the in‑vivo mouse studies, and coordinated the histopathological analyses. The Institut Pasteur provided phage isolates and expertise in phage biology, while DSMZ supplied the bacterial strains and contributed to the microbiome studies. Prof. Kai Papenfort’s group at Jena supplied transcriptomic data to dissect phage‑bacteria interactions, and the pathology team at the Freie Universität Berlin examined tissue responses.

In the microbiome‑immune interaction work package (AP 2), researchers compared the impact of broad‑spectrum β‑lactam antibiotics with the phage cocktail on the host’s microbiome‑dependent antibacterial defenses during lung infection. Antibiotic treatment disrupted the gut microbiota and reduced the production of protective metabolites, thereby weakening the lung’s innate immune response. In contrast, the phage cocktail did not impair the host’s ability to resist a subsequent *Klebsiella* pneumonia, and it markedly lowered bacterial load in a *P. aeruginosa* pneumonia model without compromising immune competence.

The phage‑efficacy and resistance package (AP 3) revealed that the success of the phage cocktail depends on the growth state of *P. aeruginosa*. Phage‑phage competition was observed, which can diminish overall lytic activity. Importantly, a synergistic combination of phage and antibiotic prevented the emergence of both phage and antibiotic resistance. Whole‑genome sequencing of phage‑resistant clones identified mutations in genes involved in lipopolysaccharide synthesis, confirming that resistance arises through alterations of the phage receptor. When the cocktail was applied to ex‑vivo cultured human lung tissue, rapid immune activation prevented the swift development of phage resistance that was seen in standard in‑vitro cultures and biofilm models.

The immunogenicity assessment (AP 4) focused on whether the phage cocktails provoke an adaptive immune response that could neutralise them. Naïve C57BL/6J mice received intraperitoneal injections of active or UV‑inactivated phage preparations. Phages were detectable in the peritoneal cavity, blood, spleen, bronchoalveolar lavage, and lung at 6 hours and remained present in the lung up to 10 days (detection limits of 83 PFU ml⁻¹ for the *P. aeruginosa* phage and 10 PFU ml⁻¹ for the *E. coli* phage). No clinical symptoms, changes in neutrophil or monocyte counts, or activation of innate or adaptive immune cells were observed. Histopathology after 21 days showed only minor germ‑center formation in the spleen, and the overall tissue architecture remained intact. These findings were published in *Viruses* (2023) and support the low immunogenicity of the phage cocktails.

Collectively, the MAPVAP project established that phage therapy can be delivered safely to the lung, that phage‑antibiotic synergy can prevent resistance, and that phage cocktails do not elicit detrimental immune responses. The collaborative framework, integrating microbiology, immunology, pathology, and transcriptomics across multiple European institutions, provides a robust foundation for advancing phage preparations toward clinical trials in ventilator‑associated pneumonia.