RETRACTED ARTICLE: Phage therapy and the microbiome in hematologic malignancies: opportunities, mechanisms, and early evidence

Abstract

Hematologic malignancies remain among the most difficult cancers to treat, challenged by profound heterogeneity, treatment-induced immune dysfunction, and the frequent emergence of drug resistance. Beyond tumor-intrinsic mechanisms, dysbiosis of the gut microbiome is increasingly recognized as a critical determinant of therapeutic outcomes, shaping hematopoiesis, immune responses, and drug metabolism. Bacteriophage (phage) therapy has re-emerged as a precision tool capable of selectively eradicating pathogenic taxa while preserving commensal short-chain fatty acid–producing communities. Preclinical and early human studies demonstrate that phages can recalibrate microbial ecosystems, disrupt antibiotic-tolerant biofilms, and enrich metabolites such as butyrate that support mucosal integrity and immune balance. Mechanistically, phage DNA enriched with CpG motifs engages Toll-like receptor 9, activating dendritic cells and enhancing cytotoxic T lymphocyte responses, suggesting dual benefits in infection control and anti-tumor immunity. Emerging applications extend further, with engineered phages serving as vectors for CRISPR–Cas gene editing, targeted cytokine delivery, and nanocarrier platforms for leukemia therapy. Despite translational promise, major hurdles persist, including immunogenicity, horizontal gene transfer, resistance evolution, and regulatory uncertainty. Addressing these challenges through GMP-compliant manufacturing, metagenomics-guided personalization, and AI-optimized cocktail design could establish phage therapy as a microbiome-informed adjunct to overcome drug resistance in blood cancers. However, direct clinical evidence of phage therapy efficacy in hematologic malignancies remains limited, and current data are largely derived from preclinical and compassionate-use contexts.

Introduction

Hematologic malignancies, including acute leukemias, lymphomas, and multiple myeloma, remain among the most challenging cancers to treat because of profound biological heterogeneity and the frequent emergence of drug resistance (Ganesan et al. 2022; Kadhum et al. 2024; Chauncey 2001). Although recent decades have witnessed substantial progress with chemotherapy, targeted therapies, and immunotherapies, many patients still experience relapse or treatment failure, underscoring the urgent need for novel strategies that can overcome resistance and sustain durable remissions (Dobosz et al. 2022; Zhong et al. 2025; Hu et al. 2025). While intrinsic genetic and epigenetic alterations of malignant cells explain part of this resistance, they do not fully account for the variability in outcomes observed across patients (Yeldag et al. 2018; Kannampuzha and Gopalakrishnan 2023; Hasan and Yang 2019). Epitranscriptomic regulation (m6A) and oncogenic long non-coding RNAs (for example CHASERR) have been implicated in tumor progression and therapy resistance in other malignancies (Wu et al. 2024). Instead, growing attention is directed toward the host microenvironment, including immune dysregulation, inflammatory signaling, and, more recently, the microbiome, as key modulators of disease behavior and therapeutic response (Hasan and Yang 2019; Ganesan et al. 2022).

The gut microbiome, often described as a hidden metabolic organ, exerts wide-ranging effects on hematopoiesis, immune regulation, and drug metabolism (Yoo et al. 2020; Okolie et al. 2025; Zheng et al. 2020). In hematologic settings, this delicate ecosystem is frequently disrupted by intensive chemotherapy, broad-spectrum antibiotics, or hematopoietic stem cell transplantation, resulting in dysbiosis (Khalil and Maher 2024; Biennier et al. 2024). Such perturbations have been linked to increased infection risk, impaired immune reconstitution, higher incidence of graft-versus-host disease (GVHD), and inferior overall survival (Guevara-Ramírez et al. 2023; Fernandez Sanchez et al. 2024). Conversely, preservation or restoration of microbial diversity, particularly the maintenance of short-chain fatty acid (SCFA)-producing taxa, correlates with improved responses to therapy, fewer infectious complications, and enhanced survival outcomes (Sun et al. 2025b).

Within this context, bacteriophages (phages), viruses that selectively infect bacteria, have re-emerged as precision tools with the potential to reshape microbial ecosystems (Luong et al. 2020, Aranaga et al. 2022). Unlike conventional antibiotics, which often indiscriminately deplete commensals, lytic phages target specific pathogenic strains while sparing beneficial microbes (Kurilovich and Geva-Zatorsky 2025). This selective activity makes phage therapy particularly attractive for immunocompromised hematology patients, in whom recurrent infections by multidrug-resistant (MDR) organisms remain a leading cause of morbidity and mortality (Mafe and Büsselberg 2025; Borysowski and Górski 2008). Beyond infection control, preclinical data suggest that phage DNA enriched with unmethylated CpG motifs can activate innate immune sensors, such as Toll-like receptor 9 (TLR9), and thereby indirectly augment anti-tumor immunity (Podlacha et al. 2024; Bollyky and Secor 2019).

Accordingly, this narrative review explores how targeted phage therapy and microbiome modulation may function as emerging adjunctive strategies against drug resistance in hematologic malignancies. We (i) outline resistance mechanisms beyond tumor-intrinsic factors, emphasizing the contribution of microenvironmental and immune–microbial dynamics; (ii) synthesize evidence on microbiome imbalance and microbial metabolites that shape hematologic outcomes; and (iii) discuss how phage-based and microbiome-informed interventions may be developed as plausible adjuncts to conventional care (Fig. 1).

Fig. 1.

Graphical summary illustrating the multifaceted roles of bacteriophage therapy in oncology: (A) microbiome modulation via targeted bacterial clearance to restore gut homeostasis; (B) enhancement of anti-cancer immunity through TLR9-mediated dendritic cell activation and cytotoxic T lymphocyte priming; (C) emerging applications in hematologic malignancies, including tumor-associated bacteria clearance, CRISPR-Cas gene delivery, and phage nanocarriers; (D) microbiome-mediated synergy with chemotherapy and infection control through fecal microbiota transplantation and phage therapy

Background

Phage therapy

Biology of phages

Bacteriophages (phages) are obligate intracellular parasites that cannot replicate outside their bacterial hosts (Naureen et al. 2020). They possess short genomes, display high host specificity, and rely extensively on host cellular machinery for replication (Strathdee et al. 2023). The most prevalent types are double-stranded DNA (dsDNA) tailed phages, consisting of a proteinaceous capsid enclosing the genome and a tail structure used for host recognition and genome injection. Infection begins when the tail binds to the bacterial cell wall, followed by injection of the genome into the cytoplasm, leaving the capsid outside (Strathdee et al. 2023).

Phages are broadly classified into lytic, lysogenic (temperate), and pseudo-lysogenic types (Zhang et al. 2022). In the lytic cycle, the phage rapidly commandeers host machinery, replicates, and ultimately lyses the host cell to release progeny, making these phages the preferred therapeutic candidates(Salmond and Fineran 2015). In the lysogenic cycle, the phage genome integrates into the host chromosome or persists as a plasmid-like episome. Under stressors such as UV irradiation or antibiotic exposure, the prophage may switch to the lytic cycle (Howard-Varona et al. 2017; Khambhati et al. 2023b). The pseudo-lysogenic cycle represents a transient state under unfavorable conditions (e.g., nutrient deprivation), in which the phage genome remains dormant without integration or replication, awaiting improved conditions before committing to lytic or lysogenic fates (Mäntynen et al. 2021, Łoś and Węgrzyn 2012; Cenens et al. 2013).

The host range, the bacterial spectrum a phage can infect, is a dynamic and adaptive trait (Marchi et al. 2023). Phages may broaden host range under selective pressure, but often at the expense of replication efficiency in novel hosts. Conversely, in environments rich in optimal hosts, phages may evolve narrower specificity. Mutations in tail fiber genes frequently underlie shifts in receptor binding, driving diversification (Holtzman et al. 2020). Beyond host range, diversity encompasses morphology (tailed, non-tailed, filamentous, enveloped), genomic material (dsDNA, ssDNA, RNA), and gene content (Liu et al. 2024b).

Phage engineering and synthetic biology

Advances in synthetic biology have enabled the precise engineering of phage genomes. Two major frameworks dominate: homologous recombination (HR) and genome rebooting. HR approaches rely on allelic exchange between the phage genome and a synthetic template during infection, though this is inefficient in strictly lytic phages (Le et al. 2013; Loessner et al. 1996). Strategies such as recombineering—using phage recombination proteins like λRed or RecE/RecT—have enhanced recombination efficiency and extended applications to Gram-positive hosts through methods like BRED. Additional refinements include reporter gene-based positive selection and CRISPR-Cas–mediated negative selection of non-recombinant progeny (Marinelli et al. 2012, Qimron et al. 2006; Hatoum-Aslan 2018).

In genome rebooting, synthetic phage genomes are assembled in vitro (e.g., Gibson assembly, transformation-associated recombination) and introduced into competent bacterial hosts for functional reconstitution (Lenneman et al. 2021). This strategy permits precise insertions, deletions, and targeted mutagenesis without interference from wild-type backgrounds or host toxicity.

Clinical rationale in the antibiotic resistance era

The global rise of MDR bacteria poses a major public health crisis, with predictions of millions of deaths by 2050. Conventional antibiotic pipelines have slowed, while resistance to even last-resort drugs continues to escalate (Lauman and Dennis 2021). escalate In this context, phage therapy, a natural, highly specific, and effective antibacterial strategy, has re-emerged as a promising alternative (Lauman and Dennis 2021; Sahu et al. 2025).

Biofilms, composed of polysaccharides, extracellular DNA, and proteins, act as protective niches that enhance bacterial tolerance to antibiotics and immune defenses (Zhao et al. 2023, Grari et al. 2025; Mayorga-Ramos et al. 2024). Phages can penetrate and disrupt biofilms either directly via lytic infection or indirectly through phage-encoded depolymerases, making them attractive for drug-resistant biofilm-associated infections (Singh et al. 2022; Urgeya et al. 2025; Islam et al. 2024). A comparative overview of the advantages of phage therapy over conventional antibiotics is presented in Table 1.

Table 1.

Comparative table: phage therapy vs. Antibiotics

Feature	Phage therapy	Antibiotics	Refs
Specificity	Highly specific; targets only specific bacterial strains	Broad-spectrum; may target both pathogens and beneficial bacteria	Dąbrowska and Abedon (2019); Cui et al. (2024)
Adaptability	Evolves alongside bacterial mutations; phages can adapt to resistance mechanisms	Limited adaptability; resistance emerges faster than new antibiotic development	Koskella and Brockhurst (2014); Subramanian (2024)
Disruption to commensals	Minimal; preserves healthy microbiota	High; often disrupts gut and skin microbiome	Torres-Barceló, (2018)
Mode of action	Infects, replicates in, and lyses specific bacterial cells	Interferes with bacterial processes (e.g., cell wall synthesis, protein production)	Lin et al. (2017)
Resistance development	Lower risk; co-evolution allows phage selection	High risk; overuse leads to multidrug resistance	Jdeed et al. (2025); Ahmed et al. (2024)
Side effects	Generally low; highly specific and self-limiting	Can cause allergic reactions, toxicity, or dysbiosis	Sawa et al. (2024)

The microbiome in hematologic health and disease

The gut microbiome represents a complex and dynamic ecosystem that contributes to nutrient absorption, drug metabolism, immune regulation, and hematopoiesis (Schubert et al. 2021; Gensollen et al. 2016). Dysbiosis, defined as a disruption of microbial homeostasis, has been linked to various hematologic disorders and malignancies (Sham et al. 2018).

A central mediator of host–microbiome interactions is the intestinal barrier, which regulates the translocation of microbial products into circulation. This multilayered defense integrates physical, chemical, immunological, and microbial components (Cui et al. 2019; Mathewson et al. 2016). Barrier dysfunction is clinically significant in settings such as graft-versus-host disease (GVHD), where severity correlates with epithelial disruption (Mathewson et al. 2016). However, its contribution to disease progression and therapy failure in hematologic malignancies such as acute myeloid leukemia (AML) remains underappreciated (Fernandez Sanchez et al. 2024).

The gut–blood axis serves as a key conduit for microbial metabolites that shape systemic immunity and hematopoiesis. Short-chain fatty acids (SCFAs), including acetate, propionate, and butyrate, derived from fiber fermentation, play essential roles in regulating the bone marrow microenvironment and systemic immune responses (Wang et al. 2024b). Enrichment of butyrate-producing taxa has been correlated with improved outcomes in multiple myeloma and CAR-T therapy (Smith et al. 2018, Shah et al. 2022), whereas depletion of these SCFA-producing microbes in conditions such as thalassemia is associated with impaired barrier integrity and increased infection risk (Nonejuie et al. 2024). Mechanistically, SCFAs modulate host metabolism through PPAR activation, histone deacetylase (HDAC) inhibition, and redox control, influencing both immune homeostasis and drug pharmacokinetics (Tsunoda et al. 2021). Emerging evidence also implicates short-chain acyl post-translational modifications as a mechanism linking microbial metabolites to oncogenic signaling and therapy response (Wu et al. 2025).

In acute lymphoblastic leukemia (ALL), dysbiosis is consistently associated with reduced SCFA levels, compromising anti-inflammatory and antitumor defenses (Song and Gyarmati 2024). Unlike synthetic HDAC inhibitors, SCFAs provide low-dose, sustained epigenetic modulation with reduced toxicity (Song and Gyarmati 2024). In AML, SCFAs induce ferroptosis and apoptosis by perturbing redox balance, increasing ROS production, and triggering mitochondrial fission/mitophagy. This ferroptotic cascade, mediated by ACSL4, leads to the release of DAMPs that activate dendritic cells and enhance antileukemic immunity (Wei et al. 2024). Additionally, SCFAs engage GPR43 (FFA2), a GPCR constitutively active in leukemic cells, to regulate cell adhesion and cytoskeletal remodeling, although not apoptosis directly (Miyasato et al. 2023). These data underscore the pivotal role of microbiome-derived metabolites in hematologic health and disease. Key gut microbiota-derived metabolites and their respective functions are summarized in Table 2.

Table 2.

Key gut Microbiota-Derived metabolites and their functions

Metabolite	Examples	Microbial source / origin	Main functions	References
Short-Chain Fatty Acids (SCFAs)	Acetate, Propionate, Butyrate	Fermentation of dietary fibers by anaerobic bacteria (e.g., Faecalibacterium, Roseburia)	Enhance Treg cell differentiation Strengthen gut barrier integrity Regulate cytokine production Support hematopoietic stem cell (HSC) function	Koh et al. (2016); Trompette et al. (2014); Belkaid and Harrison (2017)
Tryptophan Metabolites	Indole-3-acetic acid (IAA), Indole-3-aldehyde (IAld), Kynurenine	Tryptophan metabolism by gut bacteria or host enzymes	Activate AhR signaling Regulate mucosal immunity (e.g., IL-22) Influence T cell differentiation Modulate bone marrow microenvironment	Zelante et al. (2013); Agus et al. (2021); Gao et al. (2018)
Secondary Bile Acids	Deoxycholic acid (DCA), Lithocholic acid (LCA)	Transformation of primary bile acids by gut microbiota	Modulate immune responses via FXR/TGR5 Regulate inflammation May influence carcinogenesis	Hang et al. (2019); Wahlström et al. (2016; Fiorucci et al. (2020)
Polyamines	Putrescine, Spermidine, Spermine	Synthesized by gut microbes and host cells	Promote cell survival and proliferation Support immune cell differentiation Influence epigenetic regulation and HSC self-renewal	Matsumoto et al. (2012); Nowotarski et al. ( 2013); Puleston et al. (2021)
Vitamins	Vitamin K, B1, B2, B6, B12, Biotin, Folate	Synthesized by specific gut bacteria	Contribute to host vitamin pool Support hematopoiesis Participate in DNA methylation and repair mechanisms	LeBlanc et al. (2013); Rowland et al. (2018)

Bacterial taxa of clinical relevance

Beyond metabolite-mediated effects, specific bacterial taxa play a direct and clinically significant role in the pathophysiology and outcomes of hematologic malignancies. Dysbiosis in hematologic patients is often characterized by the expansion or dominance of certain pathogenic species, which compromise intestinal barrier integrity and increase susceptibility to infection, sepsis, and therapy-related complications.

Enterococcus spp., including vancomycin-resistant Enterococcus (VRE), are among the most prevalent infections in hospitalized hematology patients. In the context of intensive chemotherapy or hematopoietic stem cell transplantation (HSCT), VRE bacteremia is associated with high early mortality. In a cohort of 247 adult allogeneic HSCT patients, 27.5% exhibited pretransplant VRE colonization, with VRE accounting for 53.5% of bacteremias within the first 30 days post-transplantation and substantially increasing mortality risk (Kamboj et al. 2010). Another study confirmed that pretransplant VRE colonization independently predicted higher day-100 mortality (Zirakzadeh et al. 2008).

Enterobacteriaceae (e.g., Escherichia coli, Klebsiella spp.) are dominant causes of bacteremia in neutropenic and post-chemotherapy patients. Rapid dual-mode molecular detection platforms (for example CRISPR/Cas12a combined with RPA) permit sensitive detection of pathogenic strains and could be integrated into metagenomics-guided phage-matching workflows (Luo et al. 2024). The emergence of strains producing extended-spectrum β-lactamases (ESBLs) or carbapenemases is particularly concerning in leukemia and HSCT settings (Galloway-Peña et al. 2019; He et al. 2019). Similarly, Pseudomonas aeruginosa represents a critical opportunistic pathogen in neutropenic and post-transplant populations. Its high level of antimicrobial resistance, rapid progression to sepsis, and capacity to disrupt mucosal barriers make it a major threat to these vulnerable patients. In a large study of 1,213 cases of P. aeruginosa bacteremia in patients with hematologic malignancies, septic shock occurred in 33% of cases, with multidrug resistance (MDR/XDR), inappropriate empiric therapy, and severe neutropenia significantly increasing early mortality (Royo-Cebrecos et al. 2024).

Clostridioides difficile infection (CDI) is another frequent complication in hematology units, particularly in patients receiving chemotherapy, broad-spectrum antibiotics, or those with neutropenia. A recent study found that patients with hematologic malignancies who developed CDI had similar in-hospital mortality rates (19.5%) to noncancer patients but experienced more severe disease and higher recurrence rates. Antibiotic-induced dysbiosis, depletion of SCFA-producing commensals, VRE colonization, and infection with ribotype 078 strains have been linked to worse outcomes (Hung et al. 2021).

These bacterial taxa, Enterococcus, Enterobacteriaceae, Pseudomonas aeruginosa, and Clostridioides difficile, represent critical intersections between dysbiosis, intestinal barrier dysfunction, and clinical outcomes in hematologic malignancies. Recognizing their role not only informs risk stratification and infection prevention strategies but also highlights potential precision targets for phage therapy and microbiome-based interventions in high-risk hematology patients (Guo et al. 2024a).

In this context, bacteriophage therapy has emerged as a promising complementary approach to counter antibiotic resistance while rebalancing dysbiotic ecosystems. A study demonstrated that the phage vB_EfaS-1017 exerts potent bactericidal activity against multidrug-resistant E. faecalis, effectively inhibits and eradicates biofilms, synergizes with levofloxacin in murine bacteremia models, and promotes the recovery of beneficial genera such as Lactobacillus. This dual effect, direct bacterial killing coupled with modulation of dysbiosis, suggests that phage therapy could attenuate mucosal damage and systemic inflammation, key contributors to infectious complications and organ dysfunction in hematologic patients (Wang et al. 2025).

Clinical reports, including the PASA16 trial and several case series employing personalized phage cocktails, further demonstrate high rates of clinical improvement or microbiologic clearance in refractory infections (Onallah et al. 2023; Green et al. 2023). Mechanistically, these benefits are attributed to biofilm disruption, suppression of virulence gene expression, restoration of antibiotic sensitivity, and phage–antibiotic synergy, which may help mitigate the inflammatory and metabolic imbalances (e.g., SCFA depletion) that aggravate disease progression and treatment-related toxicity in hematologic malignancies (Green et al. 2023; Onallah et al. 2023).

Nevertheless, despite these compelling associations, direct and controlled evidence linking phage therapy to improved prognosis or modulation of disease pathways in hematologic malignancies remains limited. Most current data are extrapolated from infectious disease contexts, underscoring the urgent need for targeted clinical trials in hemato-oncologic populations to validate the therapeutic and microbiome-modulating potential of phages.

Drug-Resistant blood cancers

Classical mechanisms of multidrug resistance (MDR)

Multidrug resistance (MDR) represents one of the most formidable barriers in the treatment of hematologic malignancies. Cancer cells acquire MDR through a spectrum of mechanisms: enhanced DNA repair, reduced intracellular drug accumulation, upregulated efflux transporters, mutations of drug targets, reprogramming of apoptotic signaling, and increased drug metabolism (Yoganathan et al. 2021; Vasan et al. 2019, Sarah 2017; Krishna and Mayer 2000). This dynamic process reflects not only static genetic alterations but also a resistance continuum, wherein cancer cells adapt progressively via stress-response transcription factors such as AP1, NRF2, and ATF4 (França et al. 2024). For example, dysregulation of Wnt/β-catenin signaling (e.g., TMEM64 upregulation) has been shown to aggravate malignant phenotypes in other cancers (Yang et al. 2024).

CAR-T cell resistance in hematologic malignancies

The success of CAR-T cell therapy in B-cell leukemias and lymphomas has been tempered by both primary resistance (failure to respond) and acquired resistance (relapse after remission). Mechanistically, resistance arises through loss or downregulation of target antigens (e.g., CD19 escape), intrinsic CAR-T dysfunction leading to poor persistence, and immunosuppressive signaling within the tumor microenvironment (Bhagwat et al. 2024; Tang et al. 2024). Strategies to harness memory T-cell subsets for adoptive therapy are under active development to improve persistence and long-term tumor control in hematologic malignancies (Sun et al. 2025a). In AML, antigen specificity poses additional challenges, as targets such as CD123 are shared with normal progenitors, limiting therapeutic windows (Bhagwat et al. 2024; Tang et al. 2024). Furthermore, cytokine-driven resistance via JAK/STAT signaling exacerbates immune evasion, while toxicities such as cytokine release syndrome (CRS) complicate safe administration (Bhagwat et al. 2024; Tang et al. 2024).

Limitations of targeted therapies

While targeted therapy has shown promising results in hematologic malignancies, several limitations hinder its long-term success. Targeted therapies, particularly those aimed at oncogenes using monoclonal antibodies or small-molecule inhibitors, offer precision treatment with fewer side effects. Monoclonal antibodies remain a cornerstone of targeted therapy and, while not without risks, offer characterized safety profiles that support their use in combination regimens (Zhang et al. 2025). However, they face significant challenges in blood cancers, such as the inability to restore tumor suppressor function, difficulty distinguishing driver from passenger mutations, and the emergence of drug resistance. Moreover, some targets are considered “undruggable,” and not all patients respond equally due to genetic heterogeneity (Bishoyi et al. 2025; Zhao and Adjei 2014; Bayani et al. 2025).

Phage therapy and the microbiome: mechanisms of interaction

Modulating the microbiome to restore balance

Bacteriophages (phages) represent a highly specific biological tool capable of reshaping the gut ecosystem through the selective elimination of pathogenic or overrepresented taxa, while generally sparing commensal communities (Gordillo Altamirano and Barr 2019; Febvre et al. 2019). This selective activity, mediated largely by receptor–tail fiber interactions, distinguishes phages from broad-spectrum antibiotics and positions them as attractive candidates for microbiome engineering in vulnerable patient populations such as those with hematologic malignancies (Taslem Mourosi et al. 2022).

Evidence from human studies provides proof-of-concept for this approach. In a randomized, placebo-controlled crossover trial, phage administration significantly reduced targeted bacterial populations without altering overall microbial diversity. Importantly, this intervention was associated with an enrichment of taxa linked to beneficial metabolite production and a concomitant depletion of inflammatory-associated groups. Circulating immune mediators such as IL-4 were reduced, indicating a shift toward a more favorable host–microbial equilibrium (Febvre et al. 2019). Although these data were derived from healthy volunteers rather than cancer patients, the findings strongly suggest that phages can recalibrate the human microbiome toward a configuration enriched in SCFAs, metabolites that are mechanistically relevant to immune regulation, epithelial barrier function, and hematopoietic support.

Preclinical studies further reinforce this therapeutic logic. In disease models, phage cocktails targeting pathogenic strains reduced bacterial burden, alleviated inflammation, and preserved the integrity of commensal communities (Fig. 2) (Titécat et al. 2022). Parallel in-vitro investigations demonstrated that both whole phages and phage-derived depolymerases could dismantle antibiotic-tolerant biofilms by degrading extracellular matrices and lysing embedded bacterial populations (Amankwah et al. 2021; Hosseini Hooshiar et al. 2024). A lytic phage tested against mature biofilms achieved reductions of nearly five orders of magnitude in bacterial load, effectively dispersing biomass and illustrating the robustness of phage-driven clearance (Khalifa et al. 2015).

Fig. 2.

Microbiome modulation via bacteriophages. Bacteriophages selectively eliminate pathogenic bacteria such as Escherichia coli and Clostridium perfringens, reducing pro-inflammatory mediators like IL-4 while enriching beneficial genera such as Eubacterium that produce butyrate. This targeted clearance restores microbial balance in the gut without disrupting commensal species

The transition from these findings to hematologic contexts is compelling. Patients undergoing cytotoxic chemotherapy or hematopoietic stem cell transplantation experience profound dysbiosis, barrier dysfunction, and colonization with multidrug-resistant organisms (Soleimani Samarkhazan et al. 2025; Pötgens et al. 2023). In such settings, interventions capable of selectively suppressing inflammatory or pathogenic taxa while preserving SCFA-producing commensals could restore mucosal barrier integrity, support immune and hematopoietic recovery, and reduce infectious complications. Thus, phage therapy emerges not only as a theoretical microbiome editing tool but as a mechanistically grounded, clinically relevant strategy with direct implications for blood cancer care.

Enhancing anti‑cancer immunity via the microbiome

Beyond microbiome reshaping, phages can influence anti-cancer immunity through innate immune sensing and subsequent orchestration of adaptive responses. Their genomes are rich in unmethylated CpG motifs, which serve as ligands for Toll-like receptor 9 (TLR9) expressed by plasmacytoid dendritic cells (pDCs) and B lymphocytes (Hoshi 2025). Engagement of TLR9 triggers the canonical MyD88–IRAK–TRAF6 signaling axis, activating IRF7 and NF-κB, and culminating in the production of type I interferons and IL-12 (Kawai and Akira 2010; Honda et al. 2005). These cytokines not only enhance the activity of natural killer (NK) cells but also condition conventional dendritic cells (cDCs) to upregulate antigen presentation machinery (MHC I/II) and co-stimulatory molecules (CD80, CD86). In preclinical tumor models, such phage-conditioned cDCs migrate to draining lymph nodes, guided by CCR7 expression, and induce cytotoxic T lymphocytes (CTLs) that display robust effector programs, including granzyme B and IFN-γ secretion and improved tumor infiltration (Jiang and Zhang 2024; Alam et al. 2023, Akkaya et al. 2017).

Importantly, this immunologic cascade does not simply expand CTLs but also reprograms their functional durability. Experimental evidence suggests that the cytokine milieu and microbial signals created by phage exposure enhance T-cell metabolic capacity by promoting both glycolytic and mitochondrial pathways, increase expression of activation markers such as CD25 and CD69, and reduce exhaustion markers including PD-1 and TIM-3. Such changes produce CTLs with sustained effector function capable of overcoming checkpoint blockade resistance in otherwise refractory disease models (Huang et al. 2023) (Fig. 3).

Fig. 3.

Mechanism of anti-cancer immune enhancement via CpG-rich bacteriophage DNA stimulation of dendritic cells. CpG-rich DNA from bacteriophages is internalized by plasmacytoid dendritic cells (pDCs) and recognized within endosomes by Toll-like receptor 9 (TLR9), initiating MyD88–IRAK–TRAF6 signaling. This activates transcription factors IRF7 and NF-κB, leading to the production of type I interferons (IFN-α/β) and interleukin-12 (IL-12), promoting dendritic cell (DC) maturation, activation, and antigen presentation. Activated conventional DCs (cDCs) present tumor antigens via MHC I/II to cytotoxic T lymphocytes (CTLs), which secrete IFN-γ and granzyme B to induce tumor cell apoptosis

The immune-modulatory role of microbial metabolites further reinforces this connection. Butyrate, a prototypic short-chain fatty acid generated by gut commensals, augments CD8⁺ T-cell cytotoxicity through the GPR109A/HOPX axis, induces apoptosis in tumor cells, and has been shown to potentiate engineered T-cell therapies such as CAR-Claudin 18.2⁺ constructs by elevating IFN-γ and granzyme B output (Yu et al. 2024). High-resolution single-cell transcriptomics has additionally demonstrated that microbiome composition shapes the tumor microenvironment by shifting macrophage subsets from immunosuppressive Spp1⁺ populations toward Cd74⁺ antigen-presenting phenotypes, while simultaneously promoting γδ T-cell activation and reinforcing co-stimulatory circuits including CD40–CD40L and CD86–CD28 (Cao et al. 2025). M2 macrophage-derived exosomal long non-coding RNAs (e.g., AK083884) can metabolically reprogram macrophages and modulate inflammation in disease models (Zhang et al. 2024; Song et al. 2022). Although the majority of these findings originate from studies in solid tumors, the underlying mechanisms are broadly conserved and directly applicable to hematologic malignancies. For example, studies of the ovarian cancer microenvironment clarify how stromal and immune features modulate immunotherapy efficacy and suggest strategies potentially translatable to hematologic settings (Wang et al. 2024c). In leukemias and lymphomas, where treatment-related immunosuppression and checkpoint resistance frequently limit therapeutic success, interventions that can activate dendritic cells, rewire T-cell metabolism, and sustain cytotoxic activity are of significant translational interest. Phage therapy, by combining direct microbial modulation with immune licensing, thus emerges as a dual-action strategy that could both control infections and enhance anti-leukemic immunity.

Hematologic malignancies: disease-specific evidence, signals, and gaps

Mounting evidence demonstrates that hematologic malignancies and their therapies are closely linked to microbiome alterations (Peled et al. 2020; Wang et al. 2022). Survivors of childhood acute lymphoblastic leukemia exhibit long-term dysbiosis characterized by reduced diversity, depletion of SCFA-producing taxa, and enrichment of tryptophan-catabolizing pathways. These changes correlate with systemic inflammation and persistent immune activation, highlighting the enduring impact of therapy on host–microbial interactions (Chua et al. 2017). Comparable reductions in microbial diversity have been observed in acute myeloid leukemia, both at diagnosis and during chemotherapy, underscoring a bidirectional relationship between disease biology and microbiome disruption (Wang et al. 2022).

Mechanistic insights further clarify this connection. In AML, treatment-naïve patients display marked reductions in SCFA-producing taxa and lower fecal SCFA concentrations, along with structural disruption of epithelial tight junctions. This barrier dysfunction increases systemic translocation of bacterial components, which in turn activate anti-apoptotic programs and accelerate leukemic proliferation. Preclinical supplementation with SCFAs has been shown to restore barrier integrity, reduce systemic inflammatory burden, and diminish leukemia progression, demonstrating a direct functional link between microbial metabolites and hematologic disease (Wang et al. 2022).

Distinct microbial signatures have also been associated with other hematologic disorders. In chronic lymphocytic leukemia, dysbiotic patterns consistently emerge, characterized by relative enrichment of certain taxa and depletion of commensal groups (Faitová et al. 2022). Pediatric ALL studies using metagenomic sequencing have revealed disease-specific microbial signatures correlated with immune regulatory pathways such as IL-10 signaling, suggesting possible applications in diagnosis or risk stratification (Liu et al. 2020). The influence of therapy-related exposures adds another layer of complexity. Conditioning regimens involving ionizing radiation reshape the gut ecosystem, and certain microbiome configurations have been correlated with reduced gastrointestinal toxicity. Experimental work has shown that probiotic administration prior to irradiation can mitigate epithelial injury through innate immune signaling pathways, preserving crypt survival and reducing apoptosis (Ciorba et al. 2012). Similarly, chemotherapy alters the microbial landscape, and pre-treatment microbial composition has been identified as a predictor of febrile neutropenia and treatment-related infections (Tsunoda et al. 2021; Hakim et al. 2018).

Taken together, these hematology-specific findings underscore both the challenges and opportunities for microbiome-targeted interventions. They suggest that strategies aiming to protect or restore SCFA-producing commensals, suppress endotoxin-promoting taxa, and maintain colonization resistance could meaningfully influence disease biology and treatment outcomes. While direct clinical evidence of phage therapy in blood cancers remains absent, the convergence of mechanistic, preclinical, and associative human data provides a compelling rationale for trialing phage-guided microbiome modulation in this patient population.

Phage therapy in blood cancers: emerging application

Direct anti-tumor effects

Phage-based strategies have evolved beyond microbiome modulation to serve as platforms for gene delivery, genome editing, and therapeutic nanocarriers (Fig. 4). Preclinical studies show that engineered phages can recondition the tumor–host interface by attenuating inflammatory signaling and reshaping local microbial–immune crosstalk, effects that converge on improved barrier integrity and diminished pro-inflammatory mediators (Liu et al. 2024a; Shen et al. 2021; Eghbalpoor et al. 2024; Bullman et al. 2017). Although much of this work originates in solid tumors, the underlying mechanisms, particularly those involving immune activation and epithelial repair, are shared with hematologic settings, in which treatment-related neutropenia and transplant conditioning amplify microbiome instability (Galloway-Peña et al. 2017).

Fig. 4.

Therapeutic strategies of phage-based interventions in hematologic malignancies. Phage therapy offers multiple approaches for cancer treatment: (1) Tumor-associated bacteria clearance, bacteriophages target and eliminate Fusobacterium nucleatum within the tumor microenvironment, reducing inflammation and modulating immune responses. (2) CRISPR gene delivery – M13 phage-based vectors (phagemids) carrying CRISPR–Cas9 components deliver sequence-specific gene-editing tools into tumor cells, enabling oncogene inactivation (e.g., targeting p53) and inducing tumor cell apoptosis. (3) Phage nanocarriers – engineered M13 phages transport therapeutic genes encoding pro-apoptotic proteins, immune-activating cytokines, or tumor suppressors, enabling targeted gene delivery and effective tumor eradication

Concurrently, phages are being repurposed as therapeutic vectors through two principal strategies: (i) targeted transfer of therapeutic transgenes such as immune-stimulatory cytokines, pro-apoptotic factors, or tumor suppressors, and (ii) delivery of CRISPR–Cas systems for precise on-target editing. Foundational studies in phage-derived vectors, including phagemids and engineered temperate or virulent phages, have shown that CRISPR components can be packaged within phage capsids and delivered into mammalian cells, validating a modular framework for nucleic acid transfer (Fage et al. 2021; Zhang 2021). While classical constraints, including narrow host range, emergence of bacterial resistance, and rapid immune clearance, remain significant, advances in capsid engineering, immune-evasive designs, and delivery-route optimization (local versus systemic) are progressively reducing these translational barriers (Fage et al. 2021; Yang Zhou et al. 2020). Proof-of-concept studies illustrate the feasibility of this approach. An M13-derived bacteriophage carrying a CRISPR–Cas9 cassette against p53 successfully transduced human carcinoma cells, demonstrating that phage capsids can serve as functional gene-editing vehicles (Yang Zhou et al. 2020). Although initially tested in lung cancer models, the principle of phage-mediated delivery into malignant cells paves the way for analogous strategies in leukemic blasts, where precise genetic reprogramming could have therapeutic benefit.

Building on these foundations, advances in synthetic biology have enabled the creation of tumor-specific phage constructs capable of delivering genes encoding pro-apoptotic proteins, immune-activating cytokines such as IL-12, or tumor suppressors (Ragothaman and Yoo 2023; Wang et al. 2024a). In hematologic malignancies, such vectors are particularly attractive, as they offer selective reprogramming of leukemic cells while avoiding the systemic toxicity associated with classical viral vectors. Supporting this concept, bacteriophage-derived nanocarriers have demonstrated efficient and safe gene transfer into hematopoietic cells in preclinical leukemia models (Cui et al. 2024). For example, a minimal M13-based phage vector known as “TransPhage” has been developed to deliver therapeutic genes to human cells with high efficiency (up to ~ 95%), performing comparably or even better than adeno-associated virus (AAV) vectors. When engineered to express a membrane-bound Fc fragment, this system sensitized leukemia cells to CD16⁺ NK cell–mediated ADCC-like cytotoxicity, resulting in significant tumor reduction in xenograft models (Kao et al. 2023). In another approach, engineered P22 bacteriophage virus-like particles (VLPs) were used to encapsulate L-asparaginase, a key therapeutic enzyme in ALL. These nanoreactors exhibited improved enzyme stability and therapeutic efficacy in vitro and in vivo, offering a targeted enzyme-delivery platform that circumvents bacterial expression systems, potentially reducing immunogenicity and improving treatment outcomes in ALL (Díaz-Barriga et al. 2021). By circumventing bacterial expression systems, they reduce the risk of immunogenic impurities, providing a safer and more controllable enzyme-delivery platform for ALL therapy. Complementary approaches using plant- and phage-derived VLP shells have further diversified the landscape of asparaginase bio delivery, enhancing pharmacokinetics and reducing systemic toxicity (Tsegaye et al. 2024, Villanueva-Flores et al. 2023).

In summary, platforms such as M13 phagemids, TransPhage nanocarriers, and P22 VLP nanoreactors exemplify the translational promise of phage engineering. Together, they demonstrate how phages can be transformed from natural antibacterial agents into direct anti-tumor vectors, capable of genetic editing, immune activation, and targeted enzyme replacement. For hematologic malignancies, these strategies represent an early yet compelling frontier, uniting precision, safety, and therapeutic innovation.

Microbiome-mediated synergy with existing therapies

In hematologic malignancies, the triad of microbiome composition, anti-tumor immunity, and therapeutic efficacy is tightly interlinked. Chemotherapy, broad-spectrum antibiotics, and transplant conditioning compress microbial diversity and disrupt epithelial integrity, patterns consistently associated with impaired therapeutic response, delayed hematopoietic recovery, and increased relapse risk (Peled et al. 2020; Montassier et al. 2015). Accordingly, the gut microbiota has come to be recognized as both a key determinant of therapeutic response and a modifiable co-factor with potential to improve clinical outcomes (Fig. 5).

Fig. 5.

Synergistic role of fecal microbiota transplantation (FMT) and phage therapy in chemotherapy-associated dysbiosis. Chemotherapy can disrupt gut microbial balance, leading to dysbiosis characterized by overgrowth of pathogenic bacteria. FMT using stool from healthy donors restores microbial diversity, while bacteriophage cocktails selectively target pathogenic species. This combined approach promotes eubiosis, enhances treatment efficacy, and supports microbiome restoration during cancer therapy

Microbiome reconstitution to improve response and reduce toxicity

Mechanistic evidence indicates that commensal taxa contribute to host protection through the production of SCFAs and activation of pattern recognition receptors (PRRs) such as TLR2 and TLR5 on dendritic cells, thereby enhancing antigen presentation and T-cell priming (Gopalakrishnan et al. 2018). These pathways can synergize with cytotoxic chemotherapy agents such as cytarabine and anthracyclines, promoting apoptosis in malignant cells. Clinical translation of this concept has focused on re-compositional strategies such as fecal microbiota transplantation (FMT). A phase I trial in AML patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) demonstrated that autologous FMT accelerated microbial recovery and correlated with lower rates of GVHD and infectious complications (Taur et al. 2018). More recently, randomized early-phase studies have expanded this evidence, showing that early post-HSCT FMT can prevent severe dysbiosis and mitigate GVHD-linked inflammation (Reddi et al. 2025) and that FMT-guided microbial reconstitution supports SCFA recovery and immune homeostasis in AML/HSCT cohorts (Ebadi et al. 2025). Together, these findings align with preclinical data demonstrating that SCFA preservation and controlled PRR signaling maintain mucosal homeostasis, improve immune priming, and optimize therapeutic efficacy.

Controlling infection without exacerbating dysbiosis

Opportunistic infections remain a leading cause of morbidity and mortality in patients with hematologic cancers, particularly during neutropenia and post-transplant periods (Shono et al. 2016). Antibiotic-driven dysbiosis facilitates overgrowth of multidrug-resistant organisms (MDROs), compounding risks of sepsis, diarrhea, and secondary fungal invasion. Phage therapy has emerged as a microbiome-preserving alternative to broad-spectrum antibiotics. Compassionate-use reports in immunocompromised leukemia patients document resolution of refractory MDRO bacteremia with bespoke phage cocktails, without collateral depletion of commensal communities (Dedrick et al. 2019; Zhai et al. 2020). Beyond these early cases, clinical experience in oncology populations has shown that tailored phage therapy can control bacteremia, preserve colonization resistance, and reduce secondary fungal risk (Green et al. 2023; Mafe and Büsselberg 2025).

Challenges and future directions

Despite rapid advances, several critical barriers must be acknowledged before phage therapy can be considered a reliable adjunct in hematologic oncology. These challenges span biological, immunological, methodological, and translational dimensions. While preclinical progress has been substantial, the translation of phage therapy into hematology remains in its infancy, with only compassionate-use experiences and non-hematology oncology data available to date.

Horizontal gene transfer and biofilm dynamics

One longstanding concern is the potential of phages to mobilize undesirable genetic material. Lysogenic and pseudolysogenic states can mediate the horizontal transfer of resistance genes, toxin-encoding elements, or biofilm-associated determinants. Classic examples include lysogenic Vibrio phages transferring cholera toxin genes, and Staphylococcus aureus phages disseminating antibiotic resistance islands. In polymicrobial infections, such as those encountered in immunocompromised patients, the risk of cross-species gene transfer is amplified. Paradoxically, while many lytic phages degrade biofilms, others may inadvertently promote biofilm thickening, further complicating eradication (Doub 2021; Xu et al. 2022).

Host immune responses and pharmacokinetics

Systemic administration of bacteriophages triggers recognition as foreign antigens. Innate sensors, particularly TLR9, detect unmethylated CpG-rich phage DNA, leading to downstream NF-κB and interferon activation. These pathways accelerate clearance by phagocytes, stimulate anti-phage antibody production, and induce pro-inflammatory cytokine release. While this immunogenicity could theoretically support anti-tumor immune activation, it remains a liability in immunocompromised hematology patients. Repeated dosing risks sensitization, shorter phage half-life, and inflammatory sequelae. Mitigation strategies, encapsulation technologies, PEGylation, capsid engineering to reduce immunogenic epitopes, and localized delivery (e.g., intranasal or intratumoral), are under development but require rigorous testing in vulnerable cohorts (Zalewska-Piątek 2023; Lin et al. 2022, Guo et al. 2024b).

Bacterial resistance to phages

Much like their response to antibiotics, bacteria can quickly adapt to resist phage infection by altering surface receptors, activating CRISPR-based defenses, or triggering abortive infection programs. Because of this constant evolutionary arms race, phages must be regularly re-isolated and reformulated, which makes large-scale clinical use difficult. Researchers are therefore developing combination approaches, such as phage cocktails or phage–antibiotic synergy, to slow or prevent (Abedon 2019; Labrie et al. 2010, Pires et al. 2016).

Delivery complexity and personalization

Optimizing delivery routes remains a challenge. Oral administration faces phage inactivation in acidic gastrointestinal environments, while systemic delivery is constrained by immune clearance. Moreover, phage–microbiome interactions are highly individualized. Patient-specific microbiome variability alters adsorption, replication, and community-level impact. This makes it difficult to design “one-size-fits-all” cocktails and underscores the need for precision-matched phage–microbiome profiling using culture-based and metagenomic pipelines (Vila et al. 2024; Young et al. 2024).

Integration with CRISPR–Cas and synthetic biology

One of the most promising directions is the convergence of phages with genome-editing systems. Engineered phages can package and deliver CRISPR–Cas cassettes to disable bacterial resistance genes or virulence factors. This strategy not only expands specificity but also converts phages into programmable antimicrobials. Preclinical work has shown successful editing of bacterial genomes in situ using phage vectors (Khambhati et al. 2023a; Mayorga-Ramos et al. 2023). Extending this approach to oncology, proof-of-concept studies demonstrated that M13-derived phages carrying CRISPR–Cas9 targeting p53 could transduce malignant cells, establishing feasibility for gene-editing applications in leukemic blasts (Yang Zhou et al. 2020). However, immune activation, off-target editing, and low transduction efficiency remain barriers. Improving Cas enzyme fidelity, optimizing promoter control, and designing tumor-selective phage scaffolds are active areas of development (Fage et al. 2021; Zhang 2021).

Artificial intelligence and data-driven phage design

The rising complexity of phage–host interactions is fueling the adoption of computational tools. AI and machine learning models can analyze pathogen genomic signatures, predict phage–host compatibility, and design cocktails that maximize coverage while minimizing resistance. When combined with metagenomic sequencing, AI pipelines also permit the identification of patient-specific resistance determinants, guiding personalized phage therapy and enabling the discovery of biomarkers predictive of clinical response (Doud et al. 2025, Olawade et al. 2024; Chan et al. 2025).

Safety and regulatory imperatives

For translation into hematology, phage products must be manufactured under GMP-compliant conditions, with full genome sequencing to exclude lysogenic or virulence genes, strict endotoxin removal, sterility assurance, and reproducible batches. Given the vulnerability of neutropenic patients, contamination with bacterial debris or residual toxins carries unacceptable risks. Although purification pipelines can markedly lower endotoxin loads, many studies still lack standardized reporting of genotype, purity, or immunogenicity. Therefore, toxicology, PK/PD, and safety studies in hematology-relevant models are urgently needed before clinical adoption (Hietala et al. 2019; Liu et al. 2021).

Conclusion

Despite compelling mechanistic and preclinical signals, directly applicable clinical evidence for phage-guided microbiome modulation in hematologic malignancies remains scarce. Most human data derive either from non-hematology oncology contexts or from adjacent interventions such as FMT and observational microbiome studies in transplant and chemotherapy settings, which—while informative—do not yet establish the efficacy of phage therapy against blood cancers. Nevertheless, converging lines of evidence, linking dysbiosis to barrier dysfunction and leukemic progression, demonstrating the capacity of phages to reshape microbial communities and biofilms without broad commensal depletion, and highlighting feasible phage-based delivery platforms with immunologic co-benefits, collectively outline a credible translational path. Priorities now include rigorously designed early-phase trials in hematology cohorts, standardized GMP-grade phage manufacturing and purification, and patient-specific selection frameworks that integrate metagenomics and AI-assisted phage matching. If these hurdles are addressed, phage therapy and microbiome-informed strategies could emerge as adjuncts that mitigate infection, restore mucosal homeostasis, and help overcome drug resistance in blood cancers, warranting cautious optimism and coordinated clinical evaluation. Ultimately, while phage therapy holds strong translational promise, its clinical efficacy in hematologic malignancies awaits confirmation through well-designed prospective trials.

Author contributions

Conceptualization: Jin Liu, Alireza Bayani, literature search, data synthesis, writing, original draft, and writing, review & editing: Juanwen Zhang , Jin Liu, Alireza Bayani. The authors read and approved the final manuscript.

Funding

The author did not receive support from any organization for the submitted work.

Data availability

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethical approval

This article is a narrative review and does not contain any studies with human participants or animals performed by the author.

Consent to participate

Not applicable.

Consent to publish

Not applicable.