Prepared for NHLBI · R01 (Research Project Grant). Citations are restricted to a fixed allowed reference set (4 sources). [ILLUSTRATIVE] markers denote placeholder design parameters to be finalized with preliminary data and biostatistical input.

Resistance-Aware, Biomarker-Guided Inhaled Bacteriophage Therapy for Chronic Pseudomonas aeruginosa Infection in Non-Cystic Fibrosis Bronchiectasis

Project Summary / Abstract

Non-cystic fibrosis (non-CF) bronchiectasis is a chronic, irreversible suppurative lung disease in which permanently dilated, mucus-laden airways become a lifelong reservoir for Pseudomonas aeruginosa. Chronic P. aeruginosa infection is the dominant microbiological predictor of frequent exacerbations, accelerated lung-function decline, hospitalization, and death, yet no therapy durably reduces its airway burden. The standard of care — prolonged inhaled and systemic anti-pseudomonal antibiotics — achieves only partial, temporary suppression of antibiotic-tolerant airway biofilms while selecting for multidrug resistance, accruing toxicity, and disrupting the airway microbiome (Singh et al., 2025). Lytic bacteriophages are a mechanistically distinct option: self-amplifying, strain-specific predators that can be aerosolized directly to the infected compartment, penetrate biofilm, remain active against pan-resistant strains, and spare the broader microbiome (Singh et al., 2025).

The field has advanced from case reports to a completed, registered randomized controlled trial: Armata's Phase 2 Tailwind study (NCT05616221) of inhaled multi-phage AP-PA02 in adults with non-CF bronchiectasis and chronic pulmonary P. aeruginosa. Tailwind reported good tolerability and a reduction in sputum P. aeruginosa density that was most evident in post-hoc pooled analysis, establishing clinical plausibility while leaving the underlying pharmacology, host–pathogen dynamics, and resistance management undefined.

We propose a mechanistic, biomarker-guided R01 that converts this clinical signal into durable, generalizable knowledge. Aim 1 builds a curated, sequence-verified anti-pseudomonal phage library and validates a rapid, reproducible isolate-to-cocktail susceptibility-matching workflow against banked non-CF bronchiectasis airway isolates, with prespecified coverage and turnaround targets. Aim 2 defines, in human-sputum and biofilm models, the pharmacodynamics of phage and phage–antibiotic combinations and maps the resistance and antibiotic re-sensitization trajectories of escape mutants. Aim 3 is an eIND-enabled, single-arm pilot of nebulized personalized phage cocktails in adults with chronic pulmonary P. aeruginosa, with dense phage-kinetic, microbiological, and host-response sampling and prespecified, blinded endpoints. Together these aims address why inhaled phage works, for whom, and how to keep it working — the mechanistic questions that must be answered alongside industry efficacy trials to make inhaled phage a reliable new therapeutic class for chronic Pseudomonas airway infection, directly advancing the NHLBI mission in chronic lung disease.

Specific Aims

Chronic P. aeruginosa infection defines the highest-risk non-CF bronchiectasis phenotype, yet no therapy reliably lowers its burden without driving resistance (Singh et al., 2025). The Tailwind Phase 2 trial (NCT05616221) showed that inhaled multi-phage AP-PA02 was well tolerated and reduced sputum P. aeruginosa density — most clearly in post-hoc pooled analysis — establishing clinical plausibility but leaving the governing pharmacology, host–pathogen dynamics, and resistance management largely undefined. Critically, Tailwind included both a phage-alone cohort and a phage-plus-inhaled-antibiotic cohort but was not designed or powered to resolve whether combination therapy adds benefit. Our overarching hypothesis is that inhaled phage efficacy in this setting is governed by definable, measurable parameters — cocktail–isolate match, in-airway phage amplification ("auto-dosing"), and the fitness/re-sensitization cost of resistance — and that characterizing these parameters yields a deployable, resistance-aware treatment strategy.

Aim 1 — Build and validate a rapid isolate-to-cocktail susceptibility-matching pipeline. Hypothesis: A receptor-diverse lytic phage library plus a standardized assay can generate a patient-matched cocktail covering the majority of circulating non-CF bronchiectasis P. aeruginosa strains within a clinically actionable window. We will assemble a sequence-verified library of lytic phages spanning complementary receptors (LPS, type IV pili, flagella), quantify host range against a banked isolate collection, and lock a reproducible matching assay. Success criteria [ILLUSTRATIVE]: ≥80% of isolates covered by ≥2 complementary phages; cross-operator concordance ≥90%; turnaround suitable for Aim 3.

Aim 2 — Define phage–antibiotic pharmacodynamics and resistance trajectories in airway-relevant models. Hypothesis: Receptor-targeted cocktails and rationally chosen phage–antibiotic combinations suppress P. aeruginosa more durably than monotherapy, and phage-resistant escape mutants frequently pay fitness costs and/or re-sensitize to antibiotics. Using sputum-supplemented and biofilm systems, we will measure killing by single phages, cocktails, and cocktail-plus-inhaled-antibiotic combinations, and characterize escape mutants for fitness, virulence, receptor/efflux changes, and antibiotic susceptibility shifts. Success criteria [ILLUSTRATIVE]: ≥1 combination achieving prespecified additivity/synergy and ≥1 rotation/combination strategy that measurably delays resistance versus single-phage exposure.

Aim 3 — Conduct an eIND-enabled pilot of personalized nebulized phage cocktails. Hypothesis: Matched nebulized cocktails are feasible and well tolerated, achieve detectable in-airway phage amplification, and reduce sputum P. aeruginosa density. In a single-arm, open-label study [ILLUSTRATIVE: n≈12–15] of adults with non-CF bronchiectasis and chronic pulmonary P. aeruginosa, each participant (own pre/post control) receives an Aim 1–matched cocktail under FDA emergency/expanded-access IND oversight, with dense sampling of phage kinetics, bacterial density, emergent resistance, and host response. Success criteria [ILLUSTRATIVE]: prespecified feasibility (match generated and treatment delivered for a defined fraction of enrollees), safety consistent with prior inhaled-phage experience, and paired kinetic/density data adequate to model auto-dosing.

Impact. By linking a deployable matching pipeline, mechanistic pharmacodynamics, and a carefully monitored human pilot, this work supplies the resistance-aware, biomarker-guided foundation needed to translate a positive but early Phase 2 phage signal into a durable new therapeutic class for chronic Pseudomonas airway infection.

Significance

The problem and the NHLBI-relevant burden. Non-CF bronchiectasis is a chronic, irreversible suppurative lung disease whose natural history is dominated by P. aeruginosa. Once chronic infection is established, patients suffer more frequent exacerbations, faster lung-function decline, more hospitalizations, and increased mortality; P. aeruginosa status is the dominant microbiological prognostic marker in this population (Singh et al., 2025). This is squarely within the NHLBI mission: a progressive structural lung disease with high morbidity, recurrent acute care utilization, and no disease-modifying antimicrobial strategy.

Why current therapy fails. P. aeruginosa is intrinsically hard to clear: it forms antibiotic-tolerant biofilms in dilated airways and accumulates multidrug resistance, so prolonged inhaled and systemic anti-pseudomonal antibiotics achieve suppression at best while promoting resistance, toxicity, and airway-microbiome disruption (Singh et al., 2025). No approved therapy durably lowers P. aeruginosa burden in this setting, and the conventional antibiotic pipeline is thin.

Why phage is mechanistically matched to this niche. The target is a single, persistently colonizing pathogen in an accessible, inhalable compartment, making aerosolized delivery directly to the airway reservoir feasible (Singh et al., 2025). The clinical foundation now extends beyond anecdote. Tailwind (NCT05616221) — a completed double-blind, placebo-controlled Phase 2 RCT of inhaled AP-PA02 in adults with non-CF bronchiectasis and chronic pulmonary P. aeruginosa — demonstrated tolerability and a reduction in sputum bacterial density that was most apparent in post-hoc pooled analysis. Real-world experience is consistent: a 2025 case report describes successful induced phage-cocktail treatment of a chronic bronchiectasis patient carrying P. aeruginosa among mixed pathogens (Jernigan & Hentish, 2025), and early US expanded-access experience includes inhaled/IV anti-Pseudomonas phage (AB-PA01) in lung-transplant and bronchiectasis recipients (Aslam et al., 2020).

The gap this R01 fills. What remains missing — and what a mechanistic, NHLBI-aligned R01 is positioned to supply — is rigorous understanding of phage pharmacodynamics in human airway secretions, the determinants of response across diverse patient isolates, and the resistance-management strategies that will decide whether phage delivers durable benefit. Registration trials such as Tailwind are not designed to produce these data. Resolving them would accelerate a fundamentally new, microbiome-sparing approach to a leading driver of morbidity and mortality in chronic lung disease, and would generate transferable principles (cocktail design, auto-dosing pharmacodynamics, resistance steering) applicable across inhaled phage development.

Innovation

This proposal is innovative in four respects.

1. Reframing the question. It treats inhaled phage for bronchiectasis as a pharmacology and resistance-management problem rather than a one-off efficacy question, generating the mechanistic data that registration trials such as Tailwind (NCT05616221) are not built to produce.
2. Operationalized personalization. Rather than a fixed product, we develop and validate a rapid isolate-to-cocktail matching pipeline that selects complementary phages by receptor target to broaden coverage and raise the genetic barrier to resistance (Singh et al., 2025).
3. Resistance as an exploitable feature. We systematically map how escape mutants pay fitness costs and re-sensitize to antibiotics or lose virulence — the rational basis for phage–antibiotic combinations and cocktail rotation (Singh et al., 2025).
4. Mechanism embedded in the clinic. These readouts are embedded inside an eIND-enabled human pilot, so bacterial-density, phage-kinetic, and host-response endpoints are interpreted against the same models used to design the cocktails. The approach is deliberately matched to the current evidence base — a positive but early Phase 2 signal — and is designed to de-risk next-generation inhaled phage strategies, including depolymerase-optimized and CRISPR-Cas3–armed cocktails in development for respiratory P. aeruginosa (Singh et al., 2025).

Approach

Rigor, Reproducibility, and Authentication of Key Resources

All bacterial isolates and phage stocks will be authenticated (whole-genome sequencing of phages to confirm lytic-only lifestyle and absence of toxin/AMR genes; species/strain confirmation and susceptibility profiling of isolates) and biobanked with documented provenance. Assays will use prespecified, version-controlled protocols; in vitro experiments will be performed with biological and technical replicates and predefined statistical analysis. Clinical microbiology readouts in Aim 3 (CFU density, resistance emergence) will be performed by operators blinded to phage-kinetic results, with a prespecified analysis plan and an independent biostatistician. Sex as a biological variable will be incorporated throughout (isolate-source sex recorded; analyses examine sex-associated differences as data permit).

Aim 1 — Build and validate a rapid isolate-to-cocktail susceptibility-matching pipeline

Rationale. Phages are narrowly strain-specific, so effective products are cocktails of complementary phages ideally matched to a patient's isolate (Singh et al., 2025). A clinically deployable workflow requires both a receptor-diverse library and a fast, standardized susceptibility assay.

Experimental design. We will assemble a panel of well-characterized lytic P. aeruginosa phages chosen for complementary receptor usage (LPS, type IV pili, flagella) and confirm lytic-only lifestyle by genome sequencing. Against a biobank of non-CF bronchiectasis airway isolates [ILLUSTRATIVE: ~150 isolates], we will quantify host range, efficiency-of-plating, and quantitative liquid-culture suppression. We will lock a standardized susceptibility-testing protocol that outputs a ranked, patient-matched cocktail, and assess turnaround time and cross-operator reproducibility.

Expected outcomes. A curated, sequence-verified phage library; quantified coverage across circulating strains; and a validated matching assay with defined sensitivity and turnaround suitable for Aim 3 ([ILLUSTRATIVE] targets in Specific Aims).

Pitfalls & alternatives. If library coverage is inadequate for some isolates, we will expand the panel via additional environmental/clinical phage isolation, prioritizing broad–host-range phages. If plaque assays are variable on mucoid strains, quantitative liquid-culture kinetics will serve as the primary matching readout. If receptor classes are unevenly represented, targeted isolation against pili/flagella-dependent strains will rebalance coverage.

Aim 2 — Define phage–antibiotic pharmacodynamics and resistance trajectories in airway-relevant models

Rationale. Phages diffuse through biofilm water channels and, via depolymerases/lysins, degrade the polysaccharide matrix to reach dormant cells that antibiotics miss; sub-inhibitory antibiotics can in turn enhance phage activity, and escape mutants frequently re-sensitize to antibiotics or lose virulence (Singh et al., 2025). Because Tailwind included a phage-plus-inhaled-antibiotic cohort but was not powered to resolve combination benefit (NCT05616221), defining the mechanistic basis of combination effects is a priority.

Experimental design. In sputum-supplemented planktonic cultures and established P. aeruginosa biofilms, we will measure time-kill for single phages, matched cocktails, and cocktail-plus-inhaled-antibiotic (e.g., aminoglycoside- or polymyxin-class) combinations across clinically relevant concentrations, scoring additivity/synergy by prespecified criteria. We will isolate phage-resistant mutants and characterize growth fitness, virulence-associated phenotypes, receptor/LPS and efflux changes, and antibiotic susceptibility shifts to test the re-sensitization hypothesis. Cocktail-rotation schedules will be modeled to suppress resistance emergence.

Expected outcomes. Quantitative pharmacodynamic profiles identifying additive/synergistic phage–antibiotic combinations, and a resistance/re-sensitization map defining rotation and combination strategies for Aim 3.

Pitfalls & alternatives. If sputum components neutralize phage activity, we will titrate phage dose and test mucolytic pre-treatment. If escape mutants do not consistently re-sensitize, receptor-diverse multi-phage cocktails will be used to raise the genetic barrier to resistance, and steering toward virulence-attenuating mutations will be prioritized as the alternative objective.

Aim 3 — Conduct an eIND-enabled pilot of personalized nebulized phage cocktails

Rationale. Tailwind established that inhaled phage is tolerable and can reduce sputum P. aeruginosa density (NCT05616221); a mechanistically instrumented pilot is needed to connect in-airway phage kinetics to bacterial-density change and host response in individual patients.

Experimental design. In a single-arm, open-label feasibility study [ILLUSTRATIVE: n≈12–15] of adults with non-CF bronchiectasis and chronic pulmonary P. aeruginosa, each participant's isolate will be matched (Aim 1) to a personalized nebulized cocktail delivered via home nebulizer [ILLUSTRATIVE: twice daily for ~10 days, informed by the Tailwind treatment course]. Each participant serves as their own pre/post control. We will densely sample sputum for phage concentration (kinetics/auto-dosing), P. aeruginosa CFU density, emergent phage resistance, and host inflammatory markers, plus safety and symptom measures, with the primary microbiological readout assessed ~1 week after end of treatment [ILLUSTRATIVE: ~Day 17, mirroring Tailwind] and follow-up thereafter [ILLUSTRATIVE: through ~Day 24]. CFU and resistance assays will be run blinded to kinetic data under a prespecified analysis plan. Investigational phage will be administered under an FDA emergency/expanded-access IND (eIND) with full IRB oversight and independent DSMB monitoring.

Expected outcomes. Demonstrated feasibility of rapid matching-to-treatment; a safety/tolerability profile consistent with prior inhaled-phage experience; and paired phage-kinetic and bacterial-density data linking in-airway auto-dosing to microbiological response and on-treatment resistance.

Pitfalls & alternatives. If recruitment is slow, we will broaden inclusion across collaborating bronchiectasis clinics. If a matched cocktail cannot be generated for a candidate, the Aim 1 library will be expanded; participants without a qualifying match will not be dosed. If monotherapy responses are modest, the eIND framework permits adding an inhaled anti-pseudomonal antibiotic informed by Aim 2. A predefined interim feasibility/safety review governs continuation.

Timeline

[ILLUSTRATIVE] Years 1–2: Assemble and sequence phage library; build isolate biobank; lock and validate matching assay (Aim 1). Begin sputum/biofilm pharmacodynamics (Aim 2). Prepare eIND and IRB submissions. Years 2–4: Complete phage–antibiotic and resistance/re-sensitization studies (Aim 2); finalize regulatory approvals; manufacture and qualify clinical-grade phage stocks. Years 3–5: Enroll and treat pilot participants with dense sampling (Aim 3); integrated analysis linking in vitro pharmacodynamics to in vivo kinetics, bacterial-density change, and resistance. A go/no-go at the end of Year 2 (assay validated, ≥1 synergistic combination identified, eIND active) gates initiation of dosing.

Budget Justification (modular R01-style sketch)

[ILLUSTRATIVE] We request [ILLUSTRATIVE: $500,000] direct costs/year for [ILLUSTRATIVE: 5 years] (modular). Personnel: PD/PI (microbiology/phage biology), clinical Co-I (pulmonology), Co-I in infectious diseases/phage therapy, study coordinator, two research technicians, biostatistician, and a regulatory/quality specialist for eIND and clinical-grade stock release [ILLUSTRATIVE: ~60% of direct costs]. Phage production & characterization: sequencing, host-range and pharmacodynamic assays, endotoxin removal, and clinical-grade stock qualification [ILLUSTRATIVE: ~20%]. Clinical pilot: participant costs, nebulizers, dense sputum sampling, microbiology, phage-kinetic and host-response assays, and DSMB [ILLUSTRATIVE: ~15%]. Other: isolate biobanking, data management, publication, and dissemination [ILLUSTRATIVE: ~5%]. Costs scale with the resistance-mapping and clinical-monitoring scope rather than participant numbers.

Vertebrate Animals

Not applicable. The proposed work uses bacterial isolates, in vitro airway-relevant culture systems, and a human clinical pilot; no vertebrate animal studies are proposed. Should an inhalation-toxicology bridging study be required by the eIND, it would be added by amendment with a full Vertebrate Animal Section and IACUC approval.

Human Subjects / Clinical Trial

Aim 3 is a single-arm, open-label clinical study and constitutes an NIH-defined clinical trial. Investigational phage cocktails will be administered under an FDA emergency/expanded-access Investigational New Drug (eIND) application — the established US regulatory route for patient-specific phage products, consistent with prior US compassionate-use anti-Pseudomonas phage experience (Aslam et al., 2020). All activities will proceed under IRB approval with written informed consent and independent DSMB safety oversight; the study will be registered on ClinicalTrials.gov. Eligible participants are adults (≥18 years) with non-CF bronchiectasis and documented chronic pulmonary P. aeruginosa infection. Inclusion/exclusion criteria, prespecified stopping rules, and adverse-event reporting will follow FDA/IRB requirements. Inclusion across the lifespan, sex/gender, and race/ethnicity: enrollment will reflect the non-CF bronchiectasis population, which is enriched for older adults and women; recruitment and analysis plans address equitable inclusion, and sex will be examined as a biological variable. Dense biospecimen collection (sputum, blood) supports the mechanistic endpoints; risks are minimized by prior inhaled-phage tolerability data (NCT05616221; Jernigan & Hentish, 2025) and active monitoring. A data-sharing and resource-sharing plan will provide de-identified clinical data, phage genome sequences, and isolate/phage metadata to the community consistent with NIH policy.

Team & Environment

This program requires an integrated phage-microbiology and pulmonary-clinical team at an institution with bronchiectasis expertise, a CLIA/clinical microbiology laboratory, phage production/characterization capability, and regulatory/quality infrastructure for eIND submissions. (Named personnel and letters of support to be provided; roles below.)

• Program Director / Principal Investigator — [NAME, INSTITUTION]: phage biology, host-range and pharmacodynamic assays, overall scientific direction.
• Clinical Co-Investigator (Pulmonology) — [NAME, INSTITUTION]: non-CF bronchiectasis cohort, clinical conduct of Aim 3.
• Co-Investigator (Infectious Diseases / Phage Therapy) — [NAME, INSTITUTION]: eIND-based phage administration and compassionate-use experience (modeled on established US academic phage-therapy programs).
• Microbiology Core Lead — [NAME, INSTITUTION]: isolate biobank, susceptibility matching, resistance characterization.
• Phage Production / Quality & Regulatory — [NAME, ORGANIZATION]: clinical-grade stock qualification and eIND support (modeled on existing inhaled-phage manufacturing precedent).
• Biostatistician — [NAME, INSTITUTION]: pharmacodynamic modeling and pilot-study analysis.

Environment. Facilities will include BSL-2 microbiology, phage production/endotoxin-removal capacity, a clinical research unit for nebulized dosing and dense sampling, and institutional IRB/IND regulatory support.

References

1. Armata Pharmaceuticals, Inc. A Phase 2, Multi-Center, Double-Blind, Randomized, Placebo-Controlled Study to Evaluate the Safety, Phage Kinetics, and Efficacy of Inhaled AP-PA02 Multi-Phage Therapeutic in Subjects With Non-Cystic Fibrosis Bronchiectasis and Chronic Pulmonary Pseudomonas aeruginosa Infection (Tailwind). ClinicalTrials.gov NCT05616221; enrollment 48; completed August 2024. https://clinicaltrials.gov/study/NCT05616221
2. Singh J, Solomon M, Iredell J, Selvadurai H. Overcoming Pseudomonas aeruginosa in Chronic Suppurative Lung Disease: Prevalence, Treatment Challenges, and the Promise of Bacteriophage Therapy. Antibiotics (Basel). 2025;14(5):427. https://doi.org/10.3390/antibiotics14050427
3. Jernigan DA, Hentish RD. Successful Treatment of a Patient With Chronic Bronchiectasis Using an Induced Native Phage Cocktail: A Case Report. Cureus. 2025;17(1):e77681. https://pubmed.ncbi.nlm.nih.gov/39834667/
4. Aslam S, Lampley E, Wooten D, et al. Lessons Learned From the First 10 Consecutive Cases of Intravenous Bacteriophage Therapy to Treat Multidrug-Resistant Bacterial Infections at a Single Center in the United States. Open Forum Infect Dis. 2020;7(9):ofaa389. https://doi.org/10.1093/ofid/ofaa389