Beyond apoptosis: nanomedicine enabled reprogramming of tumor cell death for next-generation radiosensitization

Abstract

Radiotherapy remains a cornerstone in the clinical management of malignancies, leveraging DNA damage and oxidative stress to eradicate tumor cells. Nonetheless, the emergence of intrinsic and acquired radioresistance significantly compromises its therapeutic efficacy. While nanomedicine has substantially advanced radiosensitization strategies, the existing literature has largely focused on physical dose enhancement or conventional apoptosis, and the systematic reprogramming of diverse cell death modes beyond conventional apoptosis in the radiotherapy context has received less systematic attention. The present review provides a cross‑pathway synthesis of how engineered nanomaterials redirect tumor cell fate beyond apoptosis to achieve next‑generation radiosensitization, while also identifying the specific limitations and knowledge gaps that currently impede progress in this rapidly evolving field. We first delineate the hierarchical sensitization mechanisms, beginning with physical energy deposition via high-Z elements, followed by chemical amplification of reactive oxygen species through nanozyme catalysis, and biological intervention in the “6R” principles of radiobiology. Crucially, we evaluate the potential capacity of advanced nanomaterials to bypass conventional apoptotic resistance by triggering ferroptosis, pyroptosis, cuproptosis, disulfidptosis, and other emerging programmed death pathways, with a focus on the current evidence base and remaining preclinical and translational challenges. Beyond localized cytotoxicity, we highlight the mechanistic potential of nanomedicine to induce immunogenic cell death and activate the cGAS-STING pathway, suggesting a possible framework for transforming RT into an “in situ vaccine” that could reshape the immunosuppressive TME. Furthermore, we discuss the clinical translation of landmark nano-radiosensitizers, such as NBTXR3 and AGuIX, while critically addressing fundamental bottlenecks in targeting efficiency, biodistribution, and biosafety. By synthesizing current trends and future perspectives, this review contributes a strategic roadmap for advancing the design of next-generation, intelligent nanoplatforms toward more precise and systemic radiosensitization.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12951-026-04770-y.

Introduction

Given the rising incidence and mortality rates of cancer, malignant tumors remain a major challenge to global public health [1]. As a well-established and widely employed cancer treatment modality, radiotherapy (RT) plays a pivotal role in clinical oncology [1]. Clinical data indicate that more than 50% of cancer patients receive RT during their treatment [2, 3]. The primary mechanism involves the deployment of high-energy ionizing radiation to directly or indirectly damage the DNA structure of cancer cells, thereby inducing apoptosis or irreversible cell cycle arrest [4]. The lethal effect of ionizing radiation is primarily mediated by the accumulation of DNA damage, particularly double-strand breaks (DSBs) [5–7]. Additionally, RT generates abundant free radicals, predominantly reactive oxygen species (ROS), which indirectly mediate DNA damage and cell apoptosis [8]. As critical signaling molecules in oxidative stress, ROS induce cellular stress responses and apoptosis through oxidative damage to proteins, lipids, and nucleic acids [9, 10].

Despite its well‑established efficacy and broad clinical indications, the clinical reality of RT is sobering. In non‑small cell lung cancer, the 2‑year local failure rate has been reported to range from 30.7% (60 Gy) to 38.6% (74 Gy) [11, 12]; in human papillomavirus‑negative head and neck squamous cell carcinoma, the 3‑year locoregional failure rate reaches approximately 35.1% [13]; and in cT4b esophageal cancer, the 5‑year locoregional chemoradiotherapy failure rate is as high as 79.2% [14]. This quantitative evidence underscores the importance of developing new strategies to overcome radioresistance. The limited efficacy is largely attributable to tumor cell radioresistance and the complex tumor microenvironment (TME) [15–17]. Tumor cells can enhance their radiation tolerance through multiple mechanisms, including the activation of efficient DNA damage repair pathways, such as homologous recombination repair (HRR) and non-homologous end joining (NHEJ) [18, 19], cell cycle regulation [20], up-regulation of anti-apoptotic proteins [21], metabolic reprogramming [18], and the presence of cancer stem cells (CSCs) [18].

Furthermore, the physiological barriers within the TME exacerbate therapeutic resistance. Hypoxia not only attenuates ROS accumulation induced by RT but also promotes abnormal tumor vasculature and cell survival through hypoxia-inducible factor-1α (HIF-1α) induction [22–25]. Simultaneously, the phenotypic remodeling of immune cells such as regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), and M2-type macrophages, together with cancer-associated fibroblasts, dense extracellular matrix, and exosomes within the TME collectively establish a protective system. This system supports tumor tolerance and diminishes the RT response [26–30]. Radiation-induced injury further reinforces the microenvironmental barrier against cell death signals [31]. These complex biological mechanisms enable tumor cell escape from apoptosis, thereby severely limiting RT efficacy. In principle, each of these resistance‑conferring factors, including hypoxia, the immunosuppressive cellular milieu, and DNA repair proficiency, presents a rational target for nanotechnology‑based intervention, as nanocarriers can be designed to deliver radiosensitizers, oxygen, or immune modulators directly to the tumor, potentially counteracting elements that currently limit RT responses.

Radiation therapy-induced cell death primarily relies on caspase-mediated programmed apoptosis [32]. However, with deeper investigation into radiation resistance mechanisms, non-classical cell death modes, including necroptosis, pyroptosis, ferroptosis, and autophagy-related cell death, have garnered increasing attention [33, 34]. In the context of apoptotic escape, these tumor cell death pathways have been shown to mediate tumor cell clearance and promote the release of damage-associated molecular patterns (DAMPs), such as adenosine triphosphate (ATP), high-mobility group box 1 (HMGB1), and the surface exposure of calreticulin (CRT). This process, termed immunogenic cell death (ICD), has been associated with breaking immune tolerance and remodeling of the TME, thereby potentially enhancing radiosensitivity [35]. However, systematic strategies capable of simultaneously regulating multiple cell death pathways while coordinating TME intervention remain to be developed.

Nanomaterials offer promising approaches for addressing RT limitations due to their designability and multifunctionality. High-atomic-number (high-Z) nanomaterials such as Au, W, Ru, and Bi significantly enhance X-ray absorption, increase local radiation dose deposition, and amplify DNA damage [36–40]. Furthermore, certain nanomaterials catalyze ROS generation (via Fenton reactions) within the TME, potentially contributing to the overcoming of hypoxia-mediated limitations of RT [41]. Additionally, nanoplatforms can facilitate targeted therapy by conjugating specific ligands to nanoparticles for selective delivery, offering a strategy to enhance therapeutic precision while reducing harm to healthy tissues [42–44]. However, clinical translation of these nanotherapeutic strategies remains hindered by significant challenges, such as low targeting efficiency due to biological barriers within the tumor microenvironment and concerns regarding long term biosafety [42, 43]. Beyond these delivery-based approaches, a more direct strategy for overcoming radioresistance lies in the ability of nanomaterials to actively reprogram tumor cell death mechanisms. Nanomaterials can be designed to release DNA-repair inhibitors, iron ions, or small-molecule agonists in response to TME characteristics, such as acidic pH, reductive conditions, or peroxidative stress, to enhance cytotoxicity and induce specific cell death pathways. Because many of these death pathways converge on the release of DAMPs that activate innate and adaptive immunity, this capacity to directly engage cell death execution machinery also offers a route to immunomodulation. Certain functional nanoplatforms have been reported to modulate the immune microenvironment by depleting Tregs, promoting antigen presentation, or enhancing DAMPs release, suggesting potential strategies for amplifying the ICD response and facilitating synergistic effects between RT and immunotherapy.

Although several recent published review articles have discussed the development of nano-radiosensitizers, they have predominantly focused on either physical dose enhancement using high-Z materials or the modulation of a single cell death pathway. For instance, reviews by Wang et al. comprehensively catalogued high‑Z nanoparticles for physical dose enhancement [45], while works by Dong et al. summarized immunogenic cell death in nanoparticle‑mediated RT without systematically linking nanocarrier design to other cell death pathway [46]. A critical knowledge gap remains in the lack of a systematic synthesis that integrates the multi-level design of nanomaterials with the comprehensive reprogramming of diverse non-apoptotic cell death modes to overcome radioresistance. In this review, we provide a holistic overview of the current landscape of nano-radiosensitizing materials. We systematically delineate the mechanisms through which these materials modulate radiation-induced cell death pathways, focusing particularly on non-classical modes such as ferroptosis and pyroptosis as potential strategies to overcome tumor radioresistance. Furthermore, we aim to contribute to bridging the gap between bench and bedside by synthesizing clinical trial progress and the translational potential of nanomaterials, while critically evaluating the prevailing challenges and opportunities in this rapidly evolving field.

Search strategy and selection criteria

Search strategy

To ensure transparency and reproducibility, a systematic literature search was conducted in PubMed and Web of Science (WOS) to identify studies investigating nanomaterial-based strategies for modulating cell death pathways in the context of radiosensitization. The search was restricted to peer-reviewed articles published between January 2015 and September 2025. The search strategy combined three core modules: (1) Nanomedicine-related terms (e.g., nano*, nanomaterial); (2) Radiosensitization-related terms (e.g., radiosensiti*, radiotherapy enhancement, resistance reversal); and (3) Cell death mechanisms (e.g., apoptosis, ferroptosis, pyroptosis, necroptosis, cuproptosis, disulfidptosis, PANoptosis, etc.). The detailed Boolean search queries are retained in the Supplementary Methods for transparency.

Eligibility criteria and study selection

A total of 1,599 records were initially retrieved (397 from PubMed +1,202 from WOS). After the removal of duplicates, the remaining records underwent title and abstract screening against predefined eligibility criteria.

Inclusion criteria were: (i) original research articles and comprehensive reviews published in peer-reviewed journals; (ii) studies investigating the application of nanomaterials or nanoparticles in combination with radiotherapy; (iii) studies that explicitly addressed the modulation, induction, or mechanistic investigation of one or more cell death pathways in the context of radiosensitization.

Exclusion criteria included: (i) articles unrelated to radiotherapy; (ii) studies using conventional drugs without nanocarriers; (iii) conference abstracts, editorials, commentaries, book chapters, and other non-peer-reviewed materials.

After applying these criteria, a final set of 319 articles was considered eligible and included as the evidence base for this review.

Mechanisms of RT-induced cell death

Photon RT remains the clinical mainstay [47–54], generating secondary electrons (SEs) through effects like the photoelectric effect [55–58]. Nanomaterials, particularly those containing high-atomic-number (high-Z) elements, can significantly amplify the production of these low-energy electrons (LEEs), thereby contributing to the intensification of direct DNA damage and ROS generation [55] (Fig. 1).

Fig. 1.

Mechanisms of RT-induced direct and indirect damage leading to tumor cell death. X-rays induce cell death primarily through direct DNA damage (e.g., DSBs, SSBs, and BDs) caused by secondary electrons, and indirect damage mediated by ROS generated from water radiolysis. The indirect pathway leads to oxidative damage to DNA, lipids, proteins, and organelles. These combined cytotoxic effects trigger various cell death modalities, including apoptosis, autophagy, necrosis, pyroptosis, and frequently culminating in ICD

Direct DNA damage

LEEs interact with DNA to induce DSBs, single-strand breaks (SSBs), base damage (BDs), and clustered lesions [57–61]. DSBs are the most deleterious as they compromise genomic integrity [18]. Accumulated SSBs trigger replication stress and genomic instability through excessive Poly ADP-ribose polymerase (PARP) activation [62].

Cells maintain stability via repair pathways, including HRR, NHEJ, and base excision repair (BER) [63–67]. When damage exceeds repair capacity, apoptosis or senescence is triggered via the p53 pathway [68–71]. Nanotechnology-based radiosensitization strategies aim to address these repair barriers by delivering targeted inhibitors or increasing damage density to overwhelm the cellular repair machinery.

Indirect damage mediated by ROS

LEEs radiolyze water and cellular components to produce ROS, including hydroxyl radicals (•OH), hydrogen peroxide (H₂O₂), and superoxide anion (O₂•⁻) [55, 72]. As the most destructive species, •OH oxidatively modifies guanine into 8-oxoguanine (8-oxoG), leading to apurinic/apyrimidinic sites and subsequent strand breakage [58, 72–76]. While H₂O₂ and O₂•⁻ are more stable, they generate secondary •OH via iron/copper-catalyzed Fenton-like reactions [72]. Certain nanomaterials can function as nanozymes or catalysts to accelerate these reactions, potentially amplifying oxidative stress and lipid peroxidation (LPO), which may contribute to the initiation of ferroptosis [77–83]. These processes can culminate in organelle dysfunction, including mitochondrial impairment, endoplasmic reticulum (ER) stress, and lysosomal leakage [78–84].

Synergistically, direct DNA damage and ROS eliminate tumor cells while stimulating antitumor immunity through the release of DAMPs and activation of the GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway [85]. These mechanisms provide the foundation for nanocarrier-based strategies that aim to counteract radioresistance by modulating these biological pathways and remodeling the TME [18, 86].

Classification and design strategies of Nano-radiosensitizers

Nanotechnology amplifies localized energy deposition, DNA damage, and impeding cellular repair mechanisms [45, 87–90]. Physical, chemical, and biological enhancements frequently operate synergistically [45, 87–91] (Fig. 2; Table 1). While physical enhancement is robust and oxygen independent, chemical and biological strategies offer higher specificity by actively remodeling the TME as a means to address intrinsic resistance.

Fig. 2.

Mechanisms of nanomaterial-based radiosensitization combining physical, chemical, and biological strategies. Physical enhancement: Utilization of high-Z elements to enhance energy deposition via the photoelectric effect, Compton scattering, and the release of Auger electrons and secondary electrons. Chemical enhancement: Catalytic “nanozyme” activities increase intracellular ROS levels (via POD-, SOD-, and NOX-like activities) while depleting antioxidant reserves (e.g., glutathione and NADPH), leading to severe oxidative stress. Biological enhancement: Modulation of key radiobiological pathways, including inhibition of DNA repair, induction of G2/M cell cycle arrest, alleviation of hypoxia (reoxygenation), inhibition of tumor repopulation, and activation of anti-tumor immunity (e.g., through ICD induction and macrophage repolarization)

Table 1.

Comparative analysis of different nanomaterial-based radiosensitization strategies

Strategy Category	Representative Nanomaterials	Primary Mechanism	Key Advantages	Major Limitations	Key Outcomes
Physical Enhancement	AuNPs, HfO2 (NBTXR3), Bi2Te3 NPs, Pt NCs, Gd-based NPs	High-Z effect; Photoelectric and compton effects; Secondary electron emission; Auger electron cascades	Oxygen-independent; Predictable and robust dose enhancement; Clinically advanced (NBTXR3)	Lack of tumor specificity; Dose enhancement diminished at megavoltage energies; Potential long-term retention and toxicity	Massive DNA DSBs; Enhanced localized energy deposition; Increased photoelectron production
Chemical Sensitization	Fe3O4, CuO2, Mn-based nanozymes, PFCN, Au/CuNDs, TADI-COF-Fc	Catalytic ROS generation via Fenton/Fenton-like reactions; GSH depletion; Nanozyme-mimetic activities (POD, CAT, GPx, LOX, etc.)	Potent oxidative damage; TME-responsive; Capable of inducing non-apoptotic death (ferroptosis, cuproptosis)	Highly dependent on endogenous H2O2 levels and TME pH; Suboptimal enzyme specificity; Off-target redox reactions possible	Lipid peroxidation; Ferroptosis induction; GSH depletion amplifying oxidative stress; Alleviation of radio-resistance
Biological Modulation	SeD@MSNs-FA, CBL@HfO2 nanocapsules, RBCs@Se/Av, HA-PEG@CuO2	Inhibiting DNA repair (HR/NHEJ); G2/M cell cycle arrest; Anti-angiogenesis; Induction of ICD	High biological specificity; Multi-target intervention (e.g., “6R” principles); Synergy with immunotherapy	Heavy dependence on tumor genetic background; Immune-related adverse effects possible	Sustained DNA lesions; Inhibition of tumor repopulation; Reversal of immunosuppressive TME
Synergistic Platforms	Hf-BPY-Fe nMOFs, Mn3O4@GT, Hf-based nMOF	Multi-level (Physical-Chemical-Biological) integration; Passive + Active dual-targeting; Stimuli-responsive drug release	Overcomes multiple resistance mechanisms simultaneously; Enhanced tumor selectivity; Reduced systemic toxicity	Complexity in large-scale manufacturing and batch reproducibility; Long-term biodegradation and toxicity profiles insufficiently studied	Precision radiosensitization; Optimized biodistribution and clearance

NPs: Nanoparticles; AuNPs: Gold nanoparticles; HfO₂: Hafnium oxide; Bi₂Te₃: Bismuth telluride; Pt NCs: Platinum nanoclusters; Gd: Gadolinium; High Z: High atomic number; DNA: Deoxyribonucleic acid; DSBs: Double strand breaks; RT: Radiotherapy; Fe₃O₄: Iron oxide; CuO₂: Copper peroxide; PFCN: Fe-decorated CaWO₄ nanoparticles; Au/CuNDs: Gold/copper nanodots; TADI-COF-Fc: Iodine- and ferrocene-loaded covalent organic framework nanozyme; ROS: Reactive oxygen species; GSH: Glutathione; POD: Peroxidase; CAT: Catalase; GPx: Glutathione peroxidase; LOX: Lipoxygenase; TME: Tumor microenvironment; H₂O₂: Hydrogen peroxide; SeD@MSNs-FA: Folic acid-conjugated mesoporous silica nanoparticles loaded with selenadiazole derivative; CBL: Curaxin CBL0137; RBCs: red blood cells; Av: bevacizumab; HA: hyaluronic acid; PEG: polyethylene glycol; HR: homologous recombination; NHEJ: non-homologous end joining; ICD: Immunogenic cell death; Hf-BPY-Fe nMOFs: Hafnium-bipyridine-iron nanoscale metal-organic frameworks; GT: Graphene tube

Physical enhancement mechanism

High atomic number effect

RT relies on high-energy interactions with matter, primarily through Compton scattering and the photoelectric effect [92]. Compared to biological tissues, high-Z nanomaterials, such as Au [90, 93], Pt [94], Bi [95], Hf [96], Gd [97], Ta [98], W [99], Ir [100], Lu [101], Ag [102], exhibit significantly higher photon absorption cross-sections [103, 104]. These materials generate SE cascades that contribute to the induction of lethal DSBs [90, 103–105].

A critical comparison of effectiveness shows that gold nanoparticles (AuNPs, Z = 79) remain among the most extensively investigated [106, 107], with 1.9 nm particles reported to enhance RT [108] and 3.6 nm particles showing evidence of amplified energy deposition [109]. Moreover, Hf-based NBTXR3 has been advanced for clinical translation in part due to its chemical stability [110], while Pt-based systems utilize established pharmacokinetic profiles [94], and Bi-based have shown superior absorption for energy transfer in immunologically inactive tumors [95].

Geometrical parameters of nanomaterials

Nanoparticle size primarily governs biodistribution and metabolic pathways. Smaller particles tend to exhibit enhanced penetration; for instance, AuNPs (diameter < 5 nm) have been shown to achieve selective tumor accumulation and rapid clearance [111]. Similarly, Song et al. reported that smaller Bi₂Te₃ nanosheets were associated with superior tumor suppression [112]. While ultrasmall sizes may optimize biocompatibility, pH-responsive adjustable systems can prolong intratumoral retention by expanding in acidic environments [113].

Different morphologies profoundly influence surface activity and energy deposition modes [114–120]. Nanospheres generally offer dispersion [121], but complex structures like Pt nanoclusters in organic cages have demonstrated higher stability [94] and honeycomb AuNPs have been reported to maximize attenuation in deep tumors [122].

Metal-organic frameworks (MOFs) provide crystalline, porous platforms where ultrahigh porosity enables the encapsulation of diverse therapeutic cargo. MOFs offer advantages in the intrinsic integration of metal nodes for maximizing energy deposition, whereas COFs typically offer superior chemical stability owing to their purely organic covalent linkages. Utilizing this capacity, Li et al. developed a biomimetic Hf-MOF leveraging hafnium nodes for X-ray absorption and responsive nitric oxide release [123]. In contrast, Covalent organic frameworks (COFs) offer enhanced skeletal stability via organic covalent bonds, which may contribute to the secure delivery of functional moieties. For instance, Zhou et al. engineered a multifunctional COF nanozyme incorporating iodine and ferrocene that has been reported to disrupt redox homeostasis [124]. In summary, MOFs tend to offer advantages in intrinsic energy deposition through metal nodes, while COFs provide greater structural integrity that may help prevent premature payload leakage. Selecting between these frameworks requires balancing the high radiosensitization efficiency of MOF nodes against the superior cargo protection offered by COF architectures.

Chemical sensitization mechanism

Physical dose deposition is converted into cytotoxic events via chemical reactions. These mechanisms involve free radical generation via SE effects, ROS accumulation through surface catalysis, and the alleviation of hypoxia. Unlike physical dose escalation, chemical sensitization is highly dependent on TME parameters such as pH and substrate concentrations, offering targeted amplification but facing limitations in nutrient deprived regions.

Secondary electron effects

SE emission bridges physical and chemical damage. Nanoparticles promote water radiolysis, which can contribute to the damage of DNA, membranes, and proteins [125]. High-Z nanomaterials act as radiation energy amplifiers to trigger dense ionization events. LEEs (below 30 eV) generated from high-Z nanomaterials have short range [55]. This confinement creates onization hotspots and ROS bursts within tumor cells, which may enhance precision.

Electron-hole pair generation

Certain nanomaterials can generate electron-hole pairs upon excitation [126, 127]. Heterojunction architectures, such as those in Janus Ag/Ag₂S or PVP-PG nanostructures, establish internal electric fields that accelerate charge separation and suppress recombination [116, 128, 129], electrons directly damage DNA or generate O₂•⁻, while holes oxidize water to produce •OH. Furthermore, doping with selenium, has been shown to lower the electron transition energy barrier and enhance carrier density, synergistically strengthening radical generation [102]. Although these structural designs optimize charge separation, the absolute yield of ROS remains a limiting factor.

Artificial nanozymes and catalytic reactions

Certain nanozymes have been reported to regulate redox homeostasis [130]. They disrupt tumor defenses by mimicking three primary enzyme types: ROS-generating, antioxidant-depleting, and death-inducing activities.

POD-like

Iron-based nanozymes (e.g., Fe₃O₄) can serve as classical catalysts that promote apoptosis [131–133]. High-Z elements can further amplify this efficiency; for instance, HfO₂ has been shown to absorb X-rays, which can facilitate the Fe³⁺/Fe²⁺ reduction cycle [131, 132, 134]. Scintillator-based PFCN catalyze Fenton reactions, triggering ferroptosis [133]. Other transition metal-based nanoparticles, including Mn [119, 135–137], Cu [90, 138, 139], Pd [140], Pt [141], Au [142], Ir [142], W [143], also exhibit POD-like activity [144]. For example, Mn-based nanorods have been reported to amplify ROS while downregulating DNA repair proteins [135]. Similarly, Cu-based nanozymes can react with GSH to lower glutathione peroxidase 4 (GPX4) activity [90]. Notably, non-iron transition metal nanozymes (e.g., Mn, Cu) can exhibit high catalytic efficiency in acidic niches, but their overall efficacy remains limited by endogenous hydrogen peroxide levels and concerns regarding long-term systemic toxicity.

SOD-like

Nanozymes containing Cu, Zn, Fe, Mn, and nickel (Ni) frequently exhibit SOD-like activity [144]. Ferrocene-functionalized TADI-COF-Fc nanozymes have been shown to convert O₂•⁻ into H₂O₂ to fuel POD-like reactions [124]. Similarly, MnO₂@CeO_X-GAMP nanoelectrosomes can accelerate SOD-like catalysis under X-ray irradiation [145]. However, their effectiveness may be limited by the localized concentration of superoxide anions, making them less potent in tumor regions with low metabolic activity.

CAT-like

CAT-like nanozymes alleviate hypoxia by decomposing H₂O₂ into H₂O and O₂. A Cu single-atom nanozyme was reported to demonstrate catalytic efficiency exceeding that of natural catalase by 14.3 times [146]. Integrated platforms like Au/CuNDs and gadolinium-palladium nanoparticles combine CAT-like and POD-like activities to trigger cascade reactions, which can contribute to enhanced energy deposition and ROS accumulation [90, 140]. CAT-mimicry can play a meaningful role in addressing hypoxia mediated radioresistance, but its clinical impact may be restricted by the limited supply of endogenous hydrogen peroxide. Consequently, cascade designs that integrate peroxide producing enzymes may be more effective.

GPx-like

GPx-like nanozymes oxidize GSH to oxidized glutathione (GSSG), depleting antioxidant defenses to amplify oxidative stress. For example, Fe₃O₄@Bi₂S₃ nanozymes have been shown to exacerbate damage by accumulating Fe²⁺ and activating ferroptosis [147]. Similarly, FeP@Pt@HA and Au/CuNDs can disrupt redox homeostasis by converting GSH to GSSG, while disulfide-linked VGd@ICG-FA nanoprobes have been reported to release Gd ions upon GSH-triggered cleavage, which may further enhance ROS-mediated cellular destruction [90, 148, 149]. Since GSH is also vital for normal cell function, achieving high tumor-selectivity remains an important consideration for minimizing off-target systemic toxicity associated with these reactive nanozymes.

GOx-like

Metal-based nanozymes (e.g., Mn [150], Cu [151], Au [142]) mimic natural GOx via valence-state electron-transfer cycles. Specifically, the BAM nanohybrid utilizes MnO₂ for multi-enzyme cascades, while Cu₂O@Au and Ir@Au platforms deplete glucose to block energy supplies and sustain H₂O₂ levels for enhanced radiosensitization [142, 150, 151]. GOx-like activity provides a unique dual advantage by simultaneously starving radioresistant cells and fueling oxidative damage. However, the outcomes of this metabolic intervention can be heterogeneous because they are intrinsically limited by the local availability of glucose and the varied oxygen levels required for catalytic turnover across different tumor regions [152].

NDH-like

NDH-like nanozymes can disrupt redox homeostasis by catalyzing nicotinamide adenine dinucleotide phosphate (NADPH) oxidation. Among them, bismuthene-based nanozymes (BMBs) have been shown to mimic NDH activity in the acidic TME, depleting NADPH, which may contribute to the inhibition of lipid repair and the initiation of ferroptosis, thereby potentiating DNA damage [136]. This metabolic targeting specifically thwarts tumor antioxidant capacity, though its efficacy may be influenced by the baseline metabolic flux of tumors.

NOX-like

NOX-like nanozymes catalyze NADH oxidation, which can lead to the collapse of tumor metabolism. Cu-defective selenium nanozyme (Cu₂-xSe) has been shown to efficiently consume NADH under both normoxic and hypoxic conditions, inhibiting lactate production and metabolic networks. These nanozymes maintain high catalytic activity in hypoxic regions while promoting ROS accumulation through multifaceted synergistic mechanisms [153].

LOX-like

LOX-like nanozymes can directly initiate ferroptosis by catalyzing lipid peroxidation. In TADI-COF-Fc nanozymes, the ferrocene moiety has been reported to confer LOX-like activity, which may exacerbate lipid damage and contribute to radiosensitization [124]. By directly targeting membrane lipids, these nanozymes may bypass certain ROS scavenging defenses, yet their overall efficacy may be influenced by the local availability and subcellular distribution of polyunsaturated fatty acids across different tumors.

Biological enhancement mechanisms

RT efficacy is frequently curtailed by intrinsic and acquired resistance, often conceptualized through the “6R” framework of radiobiology [154, 155]. Within this biological milieu, nanomaterials have emerged as a class of radiosensitizers that can synchronize cell cycles and activate antitumor immunity, offering potential strategies to address specific aspects of the vicious cycle of radioresistance.

Targeting DNA damage and repair for radiosensitization

Nanomaterials can exacerbate DNA damage while simultaneously neutralizing repair mechanisms. High-Z nanoparticles can generate clustered DNA lesions that overwhelm cellular repair capacity [91, 95, 96, 98, 156]. For instance, CBL@HfO₂ nanocapsules have been reported to transition DNA into a radiation-sensitive Z-form, which may contribute to increased DSB intensity [157] while AuNPs downregulate critical repair genes, such as BRCA1 and RAD51, potentially delaying NHEJ and homologous recombination (HR) pathways [91, 158].

Similarly, catalytic ROS can deactivate repair-related proteins, including RAD51 [159], XRCC1 [160], 53BP1 [134, 135], and PARP [161]. For example, hollow PtCo nanospheres exhibit POD-like activity, which has been associated with suppression of the expression of repair proteins such as Ku80 and Rad51 [162]. Likewise, MnWO₄-PEG nanorods have been reported to inhibit RAD51-mediated HR and 53BP1-mediated NHEJ, which may contribute to increased radiosensitivity [135]. Although these nanoparticles can target tumor-specific vulnerabilities, the risk of collateral damage to healthy tissues remains a consideration, underscoring the importance of precise tumor delivery for therapeutic success.

Targeting cell cycle regulation for radiosensitization

Cell cycle dynamics are critical determinants of the radiation response. Since radiosensitivity peaks in the G2/M phase, nanomaterials are designed to synchronize tumor populations in sensitive phases by inducing G2/M arrest via oxidative stress or modulating cycle-related proteins. Specifically, metal oxide nanoparticles, such as UiO@MnS [119], and MnWO₄ nanorods [135], catalyze Fenton-like reactions to produce substantial •OH, which inhibits the Cyclin B1/CDK1 complex. For example, UiO@MnS has been shown to suppress CAT activity to elevate H₂O₂ levels, thereby fueling Mn²⁺-mediated Fenton reactions, and was reported to enhance RT-induced cytotoxicity by increasing the G2/M proportion from 6.6% to 15.0% [119]. Cell cycle synchronization can significantly enhance radiotherapy efficacy, although achieving uniform arrest remains challenging across heterogeneous tumor regions.

Targeting tumor repopulation for radiosensitization

Post-irradiation tumor repopulation driven by residual cell self-renewal remains a major therapeutic barrier [163, 164]. Nanomaterials can counteract this by dismantling proliferative signaling networks [165]. For instance, SeD@MSNs-FA has been reported to inhibit protein kinase B (AKT) phosphorylation while activating the c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinase (MAPK) pathways, which may contribute to thwarting repopulation [166]. Similarly, Ce6/PTX₂-NP/G@NHs nanoparticles have been shown to inactivate the phosphoinositide 3-kinase (PI3K)/AKT axis, potentially contributing to the sensitization of cancer cells [167, 168]. By concurrently blocking multiple signaling nodes, nanoplatforms like PVP-PG have shown potential to neutralize the tumor’s capacity for recovery and recurrence [116]. While multi-pathway blockade can be potent, the high degree of signaling heterogeneity across different tumor types presents challenges for the design of universal anti-repopulation nanoplatforms.

Targeting tumor oxygenation for radiosensitization

Hypoxia drives radioresistance by limiting ROS stabilization and promoting HIF-1α-mediated survival pathways. Nanomaterials can alleviate this via catalase-like activity to decompose endogenous H₂O₂ into oxygen, or through X-ray-triggered nitric oxide release to enhance blood perfusion [90, 138, 140, 146, 169]. For instance, nanoplatforms like PCN-224@IrNCs/D-Arg and scintillating nanocrystals have been reported to facilitate sustained NO release, which may enhance blood perfusion [100, 170]. Additionally, oxygen consumption can be curtailed by mitochondrial modulation. Shao et al. developed a platform that induces cuproptosis, shifting tumor metabolism from glycolysis to oxidative phosphorylation, which may contribute to reduced oxygen demand and enhance the RT response [171]. Direct oxygen generation is often substrate limited, whereas reducing metabolic oxygen demand may provide a more sustainable route to overcome hypoxia.

Targeting anti-angiogenesis for radiosensitization

Hypoxia and aberrant angiogenesis establish a self-reinforcing cycle that facilitates tumor invasion and metastasis [160]. Beyond direct cytotoxicity, nanomaterials can inhibit angiogenesis through TME modulation and targeted drug delivery [114, 160]. For instance, bio-inspired RBC-modified selenium platforms loaded with bevacizumab (RBCs@Se/Av) have been shown to generate ROS, which may impair endothelial survival and vessel repair [160], while cetuximab-targeted AuNPs have been reported to reduce CD34 expression, potentially contributing to the suppression of metastatic growth [172]. While anti-angiogenic nanotherapy can effectively starve tumors, there is a possibility that excessive vessel pruning may worsen TME hypoxia and hinder synergistic agent delivery.

Targeting cell death resistance evasion for radiosensitization

Tumor cells circumvent cell death through complex molecular mechanisms [173, 174], or antioxidant axes like solute carrier family 7 member 11 (SLC7A11)-GSH-GPX4 [175]. Nanomaterials can effectively bypass these barriers by inducing non-canonical death pathways.

Although ferroptosis represents a potent radiosensitization strategy, its efficacy may be limited by the activation of adaptive defense systems [176]. Nanomaterials can disrupt this axis by delivering inhibitors such as L-buthionine sulfoximine or RAS-selective lethal compound 3 (RSL3) [176, 177]. Furthermore, integrating GOx into MOFs has been shown to deplete NADPH, which may block cystine-to-cysteine conversion, thereby overcoming SLC7A11-mediated resistance and triggering disulfidptosis through disulfide stress [175]. While ferroptosis induction can be highly effective, its outcome may be contingent upon the metabolic flux and iron storage status of the tumor cells.

Cuproptosis offers another avenue to bypass conventional resistance [146, 178–182]. For example, HA-PEG@CuO₂ can release copper ions that bind to dihydrolipoamide S-acetyltransferase (DLAT) in the tricarboxylic acid (TCA) cycle, inducing lethal protein oligomerization [171]. Beyond copper, other metal-based systems (Ca, Zn, Na) can also initiate non-traditional death programs to enhance RT outcomes [183, 184].

Finally, ICD transforms the tumor into an “in situ vaccine” by integrating DAMPs release and microenvironment remodeling. Nanotechnology may amplify this effect by alleviating hypoxia and downregulating programmed cell death ligand 1 (PD-L1), thereby potentially contributing to the reversal of T-cell exhaustion and enabling synergy with anti-programmed cell death protein 1 (PD-1) therapy [185]. For example, core-shell structures such as HfO₂@MnO₂@GOx sustain the activation of the cGAS-STING pathway, which may help bridge localized RT with systemic antitumor immunity [186].

Targeting antitumor radioimmunity for radiosensitization

Nano-radiosensitizers have been investigated for their potential to contribute to the transition of cold tumors to hot ones by orchestrating immune remodeling [95, 100, 187]. This transition involves M2-to-M1 repolarization, Treg depletion, and enhanced infiltration of CD8 + T cells and NK cells [187–192].

Nanomaterials can amplify ICD by promoting DAMPs release. High-Z materials, such as AuNPs, have been reported to directly intensify DNA damage, which may trigger HMGB1 and ATP release, thereby promoting the maturation of dendritic cells (DCs) [193]. Furthermore, nanoplatforms like FeMn-NCe have been shown to generate an abundance of •OH, which may provoke intense ICD [194]. Additionally, multifunctional systems like C/J-Lipo^RGD integrate hypoxia alleviation with PD-L1 downregulation, which has been associated with the reversal of T-cell exhaustion, with reports of complete tumor clearance in preclinical models [185].

Several strategies have been explored to accelerate the cGAS-STING pathway. Platforms like CS-Mn have been shown to facilitate the cytoplasmic delivery of DNA fragments [177, 194, 195], while Mn²⁺ and Hf⁴⁺ ions can act as synergistic agonists that promote DC maturation and the production of type I interferons (IFNs), which may contribute to the amplification of localized and systemic RT effects [96, 196, 197].

Innovative platforms have been designed to modulate the PD-1/PD-L1 axis through mechanisms such as inhibiting transport, cleaving mRNA, or depleting cholesterol [140, 198, 199]. These integrated systems have shown potential to neutralize tumor immune evasion and may potentiate the durability of RT-induced immunity. Unlike localized death modes, ICD-based radiosensitization offers systemic benefits but remains constrained by the host’s underlying immune competence.

Synergistic antitumor strategies for radiosensitization

Tumor heterogeneity renders single-mechanism strategies insufficient, driving the trend toward multi-mechanism synergistic radiosensitization. The central objective is the integration of physical, chemical, and biological enhancements, which may contribute to improved therapeutic outcomes in a multidimensional manner. This approach aims to facilitate precise spatiotemporal control through multifunctional nanoplatforms, with the goal of enhancing overall potency and broad-spectrum adaptability across diverse oncological landscapes.

Physical-chemical-biological synergy

Intelligent nanoplatforms can amplify biological damage by integrating physical energy deposition, chemical catalysis, and biological pathway modulation [136]. High-Z elements can facilitate efficient X-ray absorption, while tailored morphologies optimize localized energy deposition. These processes trigger ROS generation or amplify oxidative stress through enzyme-mimetic reactions, which may contribute to the transformation of localized RT into systemic antitumor immunity. For instance, Hf-BPY-Fe nanoscale metal-organic frameworks (nMOFs) have been reported to achieve whole-process radiosensitization, utilizing Hf⁴⁺ for physical energy deposition; dispersed Fe³⁺/Fe²⁺ ions for in-situ Fenton reactions, and repair-protein interference for biological consolidation [134]. Similarly, the dual-atom nanoenzyme FeMn-NC_e leverages Ce for physical absorption and POD-like activity for •OH generation, with reports of activating the cGAS-STING pathway for radiodynamic immunotherapy [194]. These three-in-one synergistic systems have shown greater potency than individual strategies in preclinical studies, potentially because they can dismantle multiple radioresistance layers simultaneously. However, their multi-component complexity presents substantial hurdles for standardized pharmacokinetic modeling.

Chemical-biological synergy

Chemical-biological synergy prioritizes the cascade amplification of oxidative stress and non-canonical tumor cell death pathways, such as ferroptosis, cuproptosis, necroptosis, and ICD, over physical dose escalation [178, 194, 200–204]. For instance, the Mn₃O₄@GT nanocomposite system has been shown to amplify ROS by depleting GSH and induce necroptosis by hijacking caspase-3 activation. As an immunogenic death mode, necroptosis has been reported to sensitize tumors to PD-L1 blockade therapy, contributing to the inhibition of recurrence and the enhancement of RT outcomes [205]. These biochemical boosters can effectively bypass apoptotic resistance, but their therapeutic outcomes may be constrained by the heterogeneous metabolic flux and substrate availability across different tumor regions.

Spatiotemporal synergy and drug controlled release

Nanotherapeutic systems employ active, passive, and functional targeting as strategies to minimize toxicity. Active targeting utilizes ligands like folate (FA) for FA receptors [201], Arg-Gly-Asp (RGD) peptides for α_vβ₃ integrin [169], hyaluronic acid for CD44 [206], and cetuximab for epidermal growth factor receptors [207]. Other approaches include PD-L1 targeting [129], immune cell membrane camouflage [208], 2-deoxy-D-glucose (2DG) for metabolic targeting [209], and transferrin-based recruitment [210]. Passive accumulation relies on the enhanced permeability and retention (EPR) effect [211] while magnetic fields and bacterial carriers like Salmonella have been explored to provide additional precision [192, 212]. Emerging bio-inspired platforms, such as microorganism-assisted systems, have been investigated for their potential to offer enhanced targeting and biocompatibility. For instance, lactobacillus-coated silver nanoparticles have been reported to demonstrate selective cytotoxicity [213].

Spatiotemporal precision can be further refined by stimuli-responsive release mechanisms triggered by exogenous X-rays or endogenous TME cues. X-ray irradiation serves as a trigger that can be used to release metal ions [179, 197], carbon monoxide (CO) [214], NO [96, 161], or chemotherapy drugs [215, 216] while activating catalytic sites and photosensitizers [133, 201, 217–220]. TME-specific factors, including acidic pH, ROS enrichment, hypoxia, and high GSH or gelatinase expression, further facilitate localized cargo release. Notable examples include 8HSA8 nanoparticles for pH-responsive imiquimod (R837) delivery [202], semiconductor complexes utilizing ¹O₂-sensitive linkers for small interfering RNA (siRNA) release [202], and hypoxia-sensitive NGM platforms [221]. Additionally, GSH-triggered disulfide cleavage and gelatinase-degradable scaffolds ensure selective drug liberation at the tumor site, which may help maximize efficacy while preserving healthy tissues [206, 222]. While X-ray triggers offer advantages in spatiotemporal control and TME-responsive systems are better suited to address tumor heterogeneity; both approaches are limited by the potential for premature cargo leakage.

Multifunctional nanoplatforms for delivery and treatment

Multifunctional nanoplatforms integrate targeting, therapeutic, and imaging modules into a single system via modular assembly. Biomimetic coatings, such as cell membranes or liposomes, further endow these carriers with prolonged blood circulation and immune evasion capabilities. The convergence of radiation enhancement, drug delivery, and immune activation on a versatile nanocarrier has been investigated as a strategy to achieve multidimensional improvements in RT efficacy. For instance, Li et al. developed a biomimetic Hf-based nMOF named SHMR, which has shown potential for enhanced radio-immunotherapy [123]. Specifically, its RGD-peptide-modified red blood cell membrane (MRGD) is designed to ensure active targeting and superior biocompatibility, while the Hf nodes can maximize X-ray absorption. The encapsulation of sodium nitroprusside enables the release of NO, which may contribute to the alleviation of hypoxia and induce ICD, whereas the ligand tetrakis (4-carboxyphenyl) porphyrin (TCPP) provides fluorescence for real-time monitoring. Recent advances emphasize integrating both passive and active targeting strategies, which may help optimize clinical utility. For example, transferrin-functionalized upconversion nanoparticles (UCNP@Tf-99mTc) have been reported to achieve synergistic tumor targeting along with efficient biodistribution and renal clearance [223]. While such platforms are often developed for multimodal imaging, they provide critical insights into targeting efficiency and clearance parameters, which are essential for effective radiosensitization and minimizing systemic toxicity [223].

Nanomaterials enabled tumor cell death reprogramming for radiosensitization

The efficacy of RT is frequently undermined by intrinsic radioresistance and suboptimal systemic immune activation. Beyond traditional apoptosis, nanomaterials leverage energy deposition, ROS catalysis, and TME modulation to bypass these barriers by triggering diverse cell death pathways [224]. To provide a rigorous scientific framework, it is essential to distinguish well-established mechanisms from emerging ones. In the current landscape of cancer research, well‑established pathways (e.g., apoptosis, autophagy, pyroptosis, and ferroptosis) are supported by well‑defined molecular signatures and extensive clinical relevance. In contrast, emerging concepts (e.g., disulfidptosis and PANoptosis) offer potential novel opportunities for radiosensitization but currently require more stringent validation. The robust assessment of these pathways should rely on a multi-dimensional validation system, including: (1) morphological characterization (e.g., electron microscopy of organelle alterations or cell membrane integrity); (2) pharmacological and genetic rescue experiments to confirm causal dependency; and (3) evaluation of critical molecular nodes (e.g., LPO levels for ferroptosis, Gasdermin cleavage for pyroptosis, or actin collapse for disulfidptosis) [225]. Nanomaterials exploit the intrinsic vulnerabilities of these death pathways to overcome the radioresistance and systemic immune suppression that typically hinder RT. Beyond direct cytotoxic effects, these platforms may orchestrate a complex network of signaling cascades, ranging from metabolic collapse in cuproptosis to ionic dyshomeostasis, and can potentially culminate in ICD [226]. By integrating these diverse death modalities with precision delivery, nanomedicine reshapes the tumor microenvironment into a pro-inflammatory landscape, thereby creating a theoretical foundation for advancing RT-immunotherapy synergies (Fig. 3; Table 2).

Fig. 3.

Classification of nanomaterial-induced tumor cell death pathways for enhanced antitumor efficacy. The illustration depicts the diverse mechanisms by which various nanomaterials sensitize tumor cells to therapy. Strategies are categorized into apoptosis, autophagy, ferroptosis, cuproptosis, pyroptosis and disulfidptosis. For each modality, representative nanomaterials and key molecular biomarkers (e.g., Caspases, GPX4, FDX1, GSDM, SLC7A11) are illustrated. Furthermore, these pathways can synergistically trigger ICD, releasing DAMPs to reshape the tumor microenvironment and activate adaptive immune responses

Table 2.

Nano-radiosensitizers modulating specific programmed cell death pathways

Nanosystem	Death Pathway	Key Evidence for Cell Death	Nanomaterial Design	Sensitization Mechanisms	Key Targets	Experimental Models	Key Radiosensitization Outcomes
AHSC NPs [233]	Apoptosis	● Apoptosis: Apoptosis rate: 83.5% ↑ (4T1 cells, 6 Gy); Protein expression: Bcl-2/Bcl-xL↓, Bak↑, Bax↑, caspase-3/9↑; ROS levels↑; DNA Damage: γ-H2AX foci↑	● Core: Hf coordinated with Ce6-PEG-polyphenols ● Size: ≈21.04 nm ● Surface: Eight-arm PEG-dopamine conjugated with Ce6 ● Loading: Ato and Sab	● Physical: Hf (high-Z) ● Chemical: Ce6 (ROS) ● Biological: Ato (OCR↓, hypoxia↓); Sab (Bcl-2↓). ICB synergy	● Hf: DSB induction (γ-H2AX↑). ● Ce6: ROS generation↑ ● Ato: Complex III→ OCR↓ → HIF-1α↓ ● Sab: Bcl-2/Bcl-xL↓, Bak/Bax↑ → ↑cytochrome c release → ↑caspase-9/3 activation	● Cell models: 4T1 (breast cancer), B16F10 (melanoma), MC38 (colon adenocarcinoma) ● Animal models: BALB/c mice (subcutaneous tumor model, orthotopic and bilateral tumor model) ● Administration: AHSC NPs (dose not specified) on days 1,3,5 ● Radiation: 6 Gy X-ray, days 2,4,6	● Primary tumor volume: ≈100 mm3 (vs. >800 mm3 control, 4T1 bilateral model, 6 Gy*3) ● CD8+ Ki67+ T cells: 11.6% (vs. 3.4% control, 4T1 model) ● Apoptotic cells: 83.5% (vs. <10% control, 4T1 cells, 6 Gy) ● M2 macrophages: 6.2%↓ (vs. control, 4T1 primary tumor ● Metastatic lung nodules: markedly decreased, vs. PBS (20)
HNP NPs [96]	ICD	● Apoptosis: Late apoptosis rate: 29% ↑(CT26 cells, 2 Gy); Mitochondrial potential↓ (JC-1); γ-H2AX foci↑ ● ICD markers: CRT expression↑; ATP secretion↑	● Core: Hf4+ coordinated with 2-nIm as nanoscale coordination polymer ● Size: ≈30 nm ● Surface: Lipid bilayers containing DSPE-PEG5000 ● Loading: 2-nIm as an NO donor	● Physical: Hf (high-Z, energy deposition) ● Chemical: X-ray triggers NO release, RNS generation ● Biological: NO (DNA repair↓, hypoxia↓, apoptosis↑); Hf (cGAS-STING activation)	● Hf: DNA DSB induction (γ-H2AX↑); cGAS-STING pathway (STING↑, TBK-1↑, IRF-3↑, IFN-β↑, CXCL10↑) ● NO:2-nIm irradiation → NO release → NO + O2•− → RNS generation↑	● Cell models: CT26 (murine colon carcinoma), Hela ● Animal models: BALB/c mice, subcutaneous CT26 tumors ● Administration: Intravenous injection, 2.2 mm (100 µL), day 0 ● Radiation: 6 Gy (fractionated into 2 Gy*3 on days 1, 3, and 5)	● Tumor growth: Complete suppression (HNP + RT vs. >1000 mm3 control, CT26 s.c., 6 Gy total) ● Late apoptosis rate: ≈29% (vs. <10% control, CT26 cells, 2 Gy) ● Survival: 80% at 60 days (vs. 100% death by day 18 in controls) ● Maturation of DCs (CD86 + CD80+): ≈31.6% (vs. ≈16.7% RT alone
cGAMP/MOL [327]	ICD	● Apoptosis: Annexin-V positive↑ (MC38 cells, 2 Gy); mitochondrial potential↓ (JC-1). γ-H2AX↑ ● ICD markers:CRT exposure↑ (1.6-fold vs. PBS(+), MC38 cells, 2 Gy); ATP release↑	● Core: Hf12 SBUs and DBB-Ir photosensitizer ligands; 2D MOL ● Size: ≈190 nm ● Surface: Coordinatively bound cGAMP ● Loading: cGAMP	● Physical: Hf12 SBUs (high-Z energy deposition) ● Chemical: RT-RDT, ROS generation ● Biological: cGAMP (STING activation); ICD (CRT↑, ATP↑)	● Hf: DSB induction (γ-H2AX) ● Ir-PS: ROS generation (1O2, O2−·) ● cGAMP: STING → TBK1↑ → IRF-3↑ → IFN-β↑, CXCL10↑	● Cell models: MC38 (murine colon carcinoma), CT26 (murine colon carcinoma), Raw264.7, BMDCs, BMMs, THP-1 reporter cells ● Animal models: C57BL/6 mice (MC38 s.c./bilateral tumor models), BALB/c mice (CT26 s.c. tumor models) ● Administration: Intratumoral injection, 2 µg cGAMP/mouse, 6 consecutive days ● Radiation: 2 Gy/fraction, 6 consecutive days	● Tumor growth: CT26: 96.4% TGI (cGAMP/MOL(+) vs. PBS(+), BALB/c, 12 Gy); MC38: 99.7% TGI (cGAMP/MOL(+) vs. PBS(+), C57BL/6, 12 Gy) ● Distant tumor TGI: 70.6% (MC38 bilateral, 12 Gy)
FA-Mn3O4@ZIF-8 [201]	Apoptosis	● Apoptosis: 81.18%↑ (SiHa, 4 Gy vs. 7.58% radiotherapy alone); p-Bcl-2↓, Bax↑, cleaved caspase-3↑; Mitochondrial damage: ΔΨm collapse↓ (28.96% JC-1 monomers); DNA damage: γ-H2AX↑ (4.24-fold); ROS: 1350%↑	● Core: Mn3O4 (8 nm) ● Size: ≈200 nm ● Surface: FA-PEG modification ● Loading: None	● Physical: ZIF-8 (X-ray response ↑) ● Chemical: ROS generation; oxygenation, singlet oxygen (¹O₂) production ● Biological: GSH depletion; Tumor hypoxia relief (H2O2 decomposition); Cell cycle arrest; FA-mediated active targeting	● Mn3O4: H2O2 → O2 (hypoxia↓ → HIF-1α↓); GSH depletion → ↑ROS → DNA damage (γ-H2AX↑; XRCC1↓, p-ATM/ATR/Chk1/2 ↑) ● Apoptosis: p53 ↑, Bax ↑, caspase-3↑	● Cell models: SiHa, HeLa, C33A, HTB cervical cancer cells ● Animal models: BALB/c nude mice, SiHa subcutaneous xenograft ● Administration: 10 mg/kg, intravenous injection, bi-weekly ● Radiation: 2 Gy/fraction, every other day (total 6 fractions over 12 days)	● Tumor volume: Significant inhibition ↑ ● Apoptosis: 81.18% (vs. 7.58% RT, SiHa cells); ROS: 1350%↑ (vs. 400% RT, 4 Gy) ● HIF-1α: 42%↓ (Western blot, 15 µg/mL +4 Gy) ● EdU-positive cells: 22.81%↓ (vs. 66.68% control, 4 Gy)
Bi2S3−3BP [121]	Autophagy	● Apoptosis: TUNEL+cells↑ (vs. Bi2S3+RT, 4T1 s.c.) ● Autophagy: LC3-II/LC3-I ratio↑; P62↓ to 13.8% (4T1, Bi2S3−3BP + RT vs. Control)	● Core: Bi2S3 nanospheres ● Size: ≈260 nm ● Surface: PEG coating ● Loading: 3BP	Physical: Bi (high-Z, X-ray attenuation) Chemical: Generates ROS. Biological: 3BP (Glycolysis↓, OXPHOS↓, Oxygen consumption↓, Hypoxia relief↑, Pro-death autophagy↑)	Bi: X-ray absorption → DNA DSB induction ↑ (γ-H2AX ↑) 3BP: HK-II inhibition → glycolysis↓ → ATP↓; OXPHOS inhibition → O2↓ consumption → HIF-1α↓ Autophagy: LC3-II↑, P62↓	● Cell models: 4T1 murine breast cancer cells ● Animal models: Female BALB/c mice (4T1 subcutaneous tumor model) ● Administration: Intravenous injection, Bi2S3 5 mg/kg, 3BP 0.8 mg/kg, days 0 and 4 ● Radiation: 2 Gy/fraction, every other day (total 6 fractions over 12 days)	● Cell Viability (4T1): 11.2% (vs. Control; 2 Gy) ● Tumor growth inhibition: 79.3% (vs. Control, 4T1 s.c., day 14) ● LC3-II/LC3-I ratio: ↑
AgNPs [261]	Autophagy	● Apoptosis: Apoptosis rate ↑ (Annexin V/PI, 0.1 mM AgNPs +4 Gy) ● Autophagy: LC3-II↑ (3-fold vs. control); P62↓; Autophagic vacuoles↑	● Core:Silver rods as electrodes, PVP as stabilizer ● Size: 2 nm, 15 nm, and 40 nm ● Surface: PVP-coated ● Loading: None	● Physical: Ag (high-Z, X-ray absorption) ● Chemical: ROS generation(indirect) ● Biological: Induces DNA damage; JNK/ERK activation; Autophagy (LC3-II↑, P62↓)	● AgNPs: DNA damage (γ-H2AX↑ inferred); LC3-II conversion, P62 degradation and ERK/JNK pathways	● Cell models: U251 (human glioma), C6 (rat glioma) ● Animal models: None ● Administration: Cells incubated with AgNPs (0.05–0.2 mM, 15 nm) for 24 h before irradiation ● Radiation: 4 Gy X-ray irradiation	● Apoptosis: 11.6% (AgNPs +4 Gy vs. 2.99% 4 Gy) ● Clonogenic survival: Significantly smaller colonies
ECFC-AuNRs [115]	Ferroptosis/Autophagy inhibition	● Melanoma (Ferroptosis): GPX4↓; FTH4↓; NCOA4↑; HO-1↑ ● Breast cancer (Autophagy inhibition): LC3-II↓; Beclin-1↓; p62↓; Lautophagosomes↓	● Core: Gold nanorods (AuNRs) ● Size: ≈780 nm ● Surface: Poly (sodium 4-styrenesulfonate) (PSS) and chitosan ● Load: ECFCs act as the delivery vehicle	● Physical: Au (high-Z, X-ray absorption) ● Chemical: HT-induced ROS ● Biological: Ferroptosis (GPX4↓, HO-1↑); Autophagy inhibition (LC3-II↓, p62↓); ECFCs (Tumor-homing ↑)	● AuNRs: DNA damage (y-H2AX↑) ● Ferroptosis: GPX4↓, FTH1↓, NCOA4↑ ● Autophagy: LC3-I/II↓, Beclin-1↓, p62↓	● Cell models: M6 (melanoma), MCF-7 (breast cancer), ECFCs ● Animal models: Female athymic nude rats (RH-Foxn1rnu); Subcutaneous M6 melanoma xenografts ● Administration: Intravenous injection AuNR-enriched ECFCs (1*106 cells, day 0 and 2) ● Radiation: 2 Gy (single dose)	● Melanoma: Colony formation↓ (reduction vs. RT alone, clonogenic assay); γ-H2AX↑; GPX4↓, FTH1↓, NCOA4↑ ● Breast Cancer: Colony formation↓; γ-H2AX↑; LC3-I/II↓, Beclin-1↓, p62↓; CYTO-ID↓ ● In Vivo: Photoacoustic signal ↑28.3 units (vs. control 20.6 units, tumor mass, day 3)
N@VP [222]	Pyroptosis	● Pyroptosis: GSDMD-N↑; IL-1β↑; IL-18↑; Morphology: cell swelling and membrane blebbing ● Immune activation: HMGB1↑; ATP↑; CD80+CD86+DC cells↑; CD3+CD8+ T cells↑	● Core: VP-loaded mPEG-PVGLIG-PCL nanoparticles ● Size: ≈100 nm ● Surface: mPEG shell ● Loading: VP	● Chemical: X-ray excitation of verteporfin enhances ROS ● Biological: Pyroptosis (GSDMD-N↑); ICD (HMGB1↑, ATP↑); immune activation (DC maturation, M1 polarization, T cell infiltration)	● VP + RT: 1O2 generation↑ → ROS↑ → DNA DSB induction↑ (γ-H2AX↑) ● Pyroptosis: NLRP3↑ → Caspase-1↑ → GSDMD → GSDMD-N↑ (pore formation) → IL-1β/IL-18↑ ● Immune: Th1 cell polarization↑ (IFN-γ/IL-4↑); DC maturation↑ (CD80+/CD86+↑)	● Cell models: Lewis lung carcinoma (LLC) cells ● Animal models: C57BL/6 mice with subcutaneous LLC tumors ● Administration: Intravenous injection, 5 mg/kg (VP equivalent), days 0, 5, 11 ● Radiation:2 Gy/fraction (CDRT), 8 Gy/fraction (HDRT)	● Enhances radiotherapy: Achieving tumor inhibition comparable to high-dose with lower dose ● Induces immunomodulation: CD8+ T-cell infiltration, DC maturation
Au@AgBiS2 [269]	Pyroptosis	● Pyroptosis: Cell swelling with big bubbles; GSDME-N↑3.91-fold vs. PBS; cleaved caspase-3↑3.64-fold vs. PBS ● Apoptosis: 77.9% apoptosis rate (6 Gy) ● DAMP Release: HMGB1↑; ATP↑; CRT↑	● Core: Au nanorod core coated with AgBiS2 ● Size: ≈94 nm ● Surface: PEG-C18PMH coating ● Loading: None	● Physical: High-Z elements (Au, Bi) enhance X-ray attenuation ● Chemical: ROS burst under X-ray ● Biological: Pyroptosis induction↑	● Au, Bi: DNA DSB (γ-H2AX↑, 53BP1↑) ● Ag+: Cleaved caspase-3↑ → GSDME-N↑ ● Pyroptosis: Caspase-3↑ → GSDME cleavage↑ → GSDME-N↑ → Cell membrane pore↑ → DAMPs release↑ (HMGB1, ATP, CRT)	● Cell models: 4T1 triple-negative breast cancer cells ● Animal models: BALB/c mice (4T1 subcutaneous tumor model, metastatic lung model) ● Administration: Intravenous injection, 10 mg/kg ● Radiation: 6 Gy (single fraction, 6 h post-injection)	● Primary tumor growth inhibition: 92.29% (vs. 47.11% RT alone, 4T1 s.c., day 18) ● Lung metastatic nodules: significantly reduced vs. PBS ● Median survival: Significantly extended.vs. PBS ● Induces robust antitumor immunity: Mature DCs, effector memory T cells, CD8+T cells
DAC@O-HONs [203]	Pyroptosis	● Pyroptosis: Swollen cells with bubble-like protrusions, contents outflow; GSDME-N↑; LDH release↑; IL-18↑; Caspase-3 activity↑	● Core: Ultra-small HfO2 NPs ● Size: ≈4.5 nm ● Surface: OA-PEG ● Loading: Decitabine	● Physical: HfO2 (high-Z). ● Chemical: ROS generation ● Biological: DAC (GSDME methylation reversal↑, Pyroptosis↑)	● Hf: X-ray → ROS → DNA DSB (γ-H2AX↑) ● DAC: DNA methyltransferase inhibition → GSDME expression↑ → GSDME cleavage → pyroptosis	● Cell models:4T1 (murine TNBC) ● Animal models: BALB/c mice (4T1 subcutaneous breast cancer model) ● Administration: Intratumoral injection, 10 mg/mL (50 µL)	● Tumor growth inhibition rate: 86% (vs. 57.7% for RT only, 4T1 s.c., 6 Gy, day 20)
PWE NPs [271]	Pyroptosis	● Pyroptosis: Cell swelling and bubble formation; GSDME-N; LDH release↑; Caspase-3/cleaved caspase-3↑ ● DAMP Release:HMGB1↑	● Core: Metal-phenolic coordination between EGCG and W6+ ● Size: ≈84.5 nm ● Surface: PEGylated (mPEG-b-Pho) ● Loading: EGCG and W6+	● Physical: W6+-mediated high-Z radiosensitization ● Chemical: W6+ enhances ROS generation ● Biological: EGCG (GSDME demethylation↑, DNA repair↓); Pyroptosis induction↑	● W6+: ROS generation↑ → Caspase-3 activation↑ ● EGCG: DNMT inhibition↓ → GSDME promoter demethylation↑ → GSDME expression↑ ● Pyroptosis: Caspase-3↑ → GSDME cleavage↑ → GSDME-N↑ → Cell membrane pore↑ → DAMPs release↑	● Cell models: 4T1 (triple-negative murine breast cancer cells) ● Animal models: BALB/c mice, orthotopic 4T1, bilateral tumor, and lung metastasis models ● Administration: Intravenous injection ● Radiation: 6 Gy X-ray	● Induces pyroptosis. ● Inhibits primary & distant tumor growth ● Reduces metastasis ● Enhances CD8+ T-cell infiltration and DC maturation
Phy@PLGdH [97]	ICD	● ICD markers: HMGB1↑; ATP↑; CRT↑; CHOP expression↑; cGAS-STING activation (p-STING↑, p-IRF3↑); IFN-β↑ ● Immune markers: CD8 + T cell↑	● Core: Layered gadolinium hydroxides (LGdH) ● Size: ≈168.4 nm ● Surface: PEGylated (DSPE-PEG2000) ● Loading: Physcion	● Physical: High-Z radiosensitization ● Chemical: ROS generation ● Biological: PPP inhibition↑; NADPH/nucleotide depletione↑; DNA repair inhibition↑	● Gd: X-ray absorption↑ → ROS↑ → DNA damage (γ-H2AX↑) ● Physcion: 6PGD inhibition↓ → NADPH/ribose 5-phosphate↓ → DNA repair↓ → oxidative stress↑ ● cGAS-STING: Cytosolic DNA accumulation↑ → p-STING↑ → p-IRF3↑ → IFN-β↑	● Cell models: CT26 (murine colon carcinoma), 4T1 (murine breast cancer), Raw264.7 (macrophage) ● Animal models: BALB/c mice Subcutaneous CT26; 4T1 breast cancer metastasis model ● Administration: Intravenous injection tail vein, [Gd] = 30 mg/kg, [Phy] = 14 mg/kg ● Radiation: CT26: 5 Gy* 2 fractions (days 0 and 6); 4T1: 6 Gy* 2 fractions (days 0 and 6)	● Tumor growth inhibition: 96.8% (Phy@PLGdH + RT + αPD-L1 vs. saline, CT26 s.c) ● Systemic antitumor effects: Inhibits primary tumor growth and prevents metastasis (CT26-bearing mice and metastatic tumors on 4T1 breast tumor model) ● Induces robust ICD: Releases DAMPs, activates DC maturation and CD8+ T-cell infiltration
CS-Mn [195]	ICD	● ICD markers: cGAS-STING activation (p-STING↑, p-TBK1↑, p-IRF3↑); Cytokines in primary tumor (IFN-β↑, CXCL10↑, IL-15↑, TNF-α↑) ● Immune markers:: Enhanced tumor-specific CD8+ T-cell responses	● Core: Mn2+ mixed with chitosan ● Size: ≈591.5 nm ● Surface: None ● Loading: None	● Biological: DNA fragment delivery; cGAS-STING activation; enhanced T-cell stemness	● Mn2+: cGAS-STING activation↑ → TBK1↑ → IRF3↑ → IFN-β↑ ● DNA fragments: cytosolic accumulation↑ → cGAS-STING pathway↑	● Cell models: CT26 (colon), 4T1 (breast), DC2.4 (dendritic cells), HeLa cells ● Animal models: Female BALB/c (CT26 subcutaneous/bilateral tumor model, 4T1 orthotopic model), C57BL/6 (bilateral subcutaneous LLC). Bilateral CT26; orthotopic 4T1 ● Administration: Intratumoral injection, dose:13.75 µg (CT26 model)/13.75*2 µg (LLC model) Mn²⁺ per mouse ● Radiation: CT26: 5 Gy*1–5 Gy*2 (days 0 and 8); LLC: 5 Gy*2 (days 0 and 8); 4T1: 6 Gy* 2 (days 0 and 8)	● Inhibited the primary tumor, distant tumors and even systemic metastases ● Enhance immunity (Mature DC cells and CD8+ T cells)
Th-Ir-DBB/Dig [199]	Disulfidptosis/ICD	● Disulfidptosis: NADP+/NADPH ratio↑; GSSG/GSH ratio↑; intracellular cystine↑; Late apoptotic 4T1 cells (39.7%, BAY-876 + Th-Ir-DBB/Dig vs. 24.7% Th-Ir-DBB/Dig alone); F-actin contraction↑ ● ICD markers: Upregulation of ROS, γ-H2AX, and release of DAMPs (CRT↑, ATP↑, HMGB-1↑)	● Core: Th6O4(OH)4 SBUs and Ir-DBB ligands ● Size: ≈106 nm ● Surface: None ● Loading: Digitonin	● Physical: High-Z radiosensitization (Th) ● Chemical: Radiodynamic therapy (Ir-DBB) (ROS↑) ● Biological: Digitonin (Cholesterol↓, GLUT1↓, Glucose↓, GSH↓, Disulfidptosis↑)	● Th: DSB induction (γ-H2AX↑) ● Ir-DBB: 1O2 generation↑/·OH generation↑ ● Digitonin: Cholesterol depletion↑ → GLUT1 downregulation↓ → Glucose uptake↓ → NADPH/GSH depletion↓ → Disulfidptosis↑	● Cell models: 4T1 (breast cancer), CT26 (colon cancer), LLC (lung cancer) ● Animal models: BALB/c mice, subcutaneous 4T1 and CT26 ● Administration: Intratumoral injection, Th dose 0.5 µmol, Digitonin dose 0.19 µmol ● Radiation: 2 Gy/fraction, 3 fractions, total 6 Gy	● CT26 tumor growth inhibition (TGI): 97.7% (Th-Ir-DBB/Dig + RT vs. PBS, 2 Gy * 3) ● 4T1 tumor TGI: 86.1% (Th-Ir-DBB/Dig + RT vs. PBS, 2 Gy * 3) ● Enhances immunogenicity (4T1 model): Cd47-positive 4T1 cells↓; Mature DC cells↑; CD3+CD4+ helper T cells↑; CD3+CD8+ helper T cells↑; IL-6, TNF-α, IFN-γ↑.
GOx@FBSP MOFs [175]	Disulfidptosis/Ferroptosis/ICD	● Ferroptosis: Increased ROS, lipid peroxidation (BODIPY C11) ↑, PTGS2↑ and GPX4↓ ● Disulfidptosis: Increased intracellular cystine and altered actin cytoskeleton; NADP+/NADPH ratio↑ ● ICD: DAMPs (CRT↑, ATP↑, HMGB-1↑)	● Core: Fe and Bi dual ions bridged by disulfide bonds (FBSP) ● Size: ≈199.64 nm. ● Surface: PEGylated ● Loading: Glucose oxidase (GOx)	● Physical: High-Z elements enhance radiation energy deposition ● Chemical: Fenton reaction (ROS generation↑), GOx (glucose depletion↑) ● Biological: GSH depletion↑, Ferroptosis↑, Disulfidptosis↑, ICD↑	● Fe: Fenton reaction → ·OH ● Bi: high-Z radiosensitization ● Disulfide bonds: GSH depletion → GPX4↓, PTGS2↑ ● GOx: glucose → gluconic acid + H2O2 → pH↓, OH↑; glucose depletion → NADPH↓ → cystine↑ → disulfide stress → F-actin collapse	● Cell models: 4T1 (breast cancer), MC3T3-E1 (normal cells) ● Animal models: BALB/c mice (4T1 subcutaneous tumor model) ● Administration: Intravenous tail vein injection, single dose, 5 mg/mL ● Radiation: 6 Gy (single dose)	● Reduces tumor growth: The tumor volume was significantly reduced in GOx@FBSP + RT group; TUNEL-positive cells↑ ● Enhances immunogenicity: Mature DC cells↑; CD8+ T cells↑
Fe@Pdots-siRNA [280]	Ferroptosis	● Ferroptosis: ROS↑ and lipid peroxidation↑; GSH depletion; SLC7A11↓; GPX4↓; Fer-1 reverses the cytotoxic effect of Fe@Pdots-siRNA + RT; PTGS2 expressions↑	● Core: Iron-doped semiconducting polymer nanoparticles (MEH-PPV core with Fe2+) ● Size: ≈140 nm ● Surface: NPC CNE2 cell membrane ● Loading: siRNA targeting circADARB1	● Chemical: Fe2+ amplifies Fenton reaction ● Biological: siRNA knocks down circADABRB1 to inhibit HSP90B1/SLC7A11/GPX4 axis	● Fe2+/Fenton: H2O2 → ·OH↑ → Lipid peroxidation↑ ● circADARB1 (Targeted by siRNA): HSP90B1 expression↓ → SLC7A11/GPX4↓ stability → Ferroptosis sensitivity↑	● Cell models: CNE2, HONE1, 5–8 F (human NPC cell lines) ● Animal models: BALB/c nude mice (CNE2 subcutaneous xenograft model) ● Administration: Intravenous injection (10 mg/kg) ● Radiation: 6 Gy (single dose)	● Reduces tumor growth: The tumor volume was significantly reduced in Fe@Pdots-siRNA group
Fe-doped Mn3O4 [89]	Ferroptosis	● Ferroptosis: Lipid peroxidation (4-HNE↑); GPX4↓; STING pathway activation	● Core: Fe-doped Mn3O4 ● Size: ≈20 nm ● Surface: dp-PEG-mal and SH-HA ● Loading: None	● Chemical: Fenton reaction (ROS generation↑) ● Biological: Ferroptosis↑; STING pathway activation↑	● Fe2+: Lipid peroxidation (4-HNE↑) → GPX4↓ → ferroptosis ● Mn2+: cGAS-STING pathway (STING↑, IFN-β↑) → immune activation (MDSC↓, CD8+↑, DC maturation↑)	● Cell models: 4T1 (murine breast cancer), RR-4T1 (radioresistant 4T1) ● Animal models: Subcutaneous inoculation of RR-4T1 cells in BALB/c mice ● Administration: Intratumoral injection, dose 2 mg kg− 1 ● Radiation: 8 Gy X-ray, single dose per irradiation day, given 6 h after injection, on days 0, 3, 6 (total 3 fractions)	● Reduces tumor growth: The tumor volume was significantly reduced in Mn3O4:15%Fe + IR group ● Reduces metastasis: Significantly inhibits lung metastasis, factor of 4.5 (vs. IR group) ● Enhances immunogenicity: Mature DC cells↑; CD8+ T cells↑
PFCN [133]	Ferroptosis/ ICD	● Ferroptosis: Lipid peroxidation (4-HNE↑); GPX4↓; GSH↓; MDA↑; Mitochondrial atrophy, fragmentation, membrane potential ● ICD: DAMPs (CRT↑, ATP↑, HMGB-1↑)	● Core: CaWO4 core (nanoscintillator); Fe3O4 surface decoration ● Size: ≈86 nm ● Surface: PEG (600) coating ● Loading: None	● Physical: X-ray-triggered UV emission; High-Z ● Chemical: Fe3+ to Fe2+ photoreduction; Fenton reaction ● Biological: Ferroptosis (GPX4↓, LPO↑, GSH↓); ICD (CRT, ATP, HMGB1)	● CaWO₄: X-ray → UV light → Fe3+ photoreduction ● Fe2+: Fenton reaction (·OH) → lipid peroxidation → GPX4↓, GSH↓, MDA↑	● Cell models: 4T1 breast cancer cells; 3T3 normal cells ● Animal models: Female BALB/c mice; Subcutaneous tumor model (4T1 cells) ● Administration: Intravenous injection (5 mg/kg Fe equivalent) ● Radiation: 6 Gy (in vitro/in vivo therapy) or 12 Gy (in vitro mechanistic studies)	● Tumor growth inhibition rate: 96.55% (vs. PBS, 4T1 s.c., 6 Gy) ● Tumor volume:The tumor volume was significantly reduced in PFCN + X-ray group ● Enhances immunogenicity: Mature DC cells↑; CD8+ T cells↑
PWCu [178]	Cuproptosis/ ICD	● Cuproptosis: FDX1 and LIAS were upregulated in post-radiotherapy clinical samples and radioresistant models; ● ICD: DAMPs (CRT↑, HMGB-1↑)	● Core: Two polyoxoanion units encapsulating four coplanar CuO6 octahedra for ultra-high copper exposure ● Size: Ultra-small (1–1.6 nm) ● Surface: None ● Loading: None	● Physical: high-Z, energy deposition ● Chemical: Reduction (Cu2+ → Cu+) ● Biological: Cuproptosis (protein lipoylation aggregation↑, proteotoxic stress↑). ICD (DAMPs↑, immune activation↑)	● PWCu: X-ray → Cu2+ reduction ● Cu+: binds lipoylated DLAT/DLST → proteotoxic stress; destabilizes Fe-S cluster proteins (POLD1, ACO2) → TCA cycle disruption; HSP70↑, CRT exposure, HMGB1 release	● Cell models: 4T1, 4T1-R, B16-F10, B16-F10-R, A549, A549-R, HCT116, HCT116-R, 143B, MDA-MB-231, HeLa, Hepa1-6, CT26, BEAS-2B, NCM-460, HK-2, AC16, HaCaT, HUVEC ● Animal models: BALB/c and C57BL/J mice. Tumor models: Bilateral flank subcutaneous (4T1-R), Orthotopic breast (4T1) ● Administration: Intratumoral injection, 500 µg, every 2 days for 3 fractions ● Radiation: 2 Gy per fraction, 3 fractions (daily or every 2 days). Primary irradiation for recurrent model: 8 Gy*4. Re-irradiation: 2 Gy*3	● Complete response: 40% (both radioresistant and re-irradiation tumour models) ● Tumor volume: Near-complete elimination of primary tumors ● Median survival: 90 days (PWCu + IR) vs. 56 days (PWCu or IR alone) ● Lung metastases: Number of metastatic nodules↓ ● Enhances immunogenicity: CD8+ T cells↑
HA-PEG@CuO2[171]	Cuproptosis	● Cuproptosis: DLAT oligomer↑, FDX1↓, LIAS↓; GSH/GSSG ratio↓ ● Metabolic Reprogramming: ATP levels↓ (SMMC 7721, HA-PEG@CuO₂ + X-ray); Lactate levels↓ (SMMC 7721, HA-PEG@CuO₂ + X-ray); HIF-1α protein expression↓ (Western blot, HA-PEG@CuO₂)	● Core: PVP@CuO2 ● Size: ≈162.9 nm ● Surface: PEG-SS-PEG and and HA ● Loading: 2-nIm as an NO donor	● Chemical: Fenton-like reaction (ROS generation, GSH depletion) ● Biological: Cuproptosis induction; Metabolic reprogramming (OXPHOS promotion)	● CuO₂: Cu2+ → Cu+ (redox cycle). ● Fenton-like: H2O2 + Cu+ → ·OH↑ ● Cuproptosis: FDX1↓, LIAS↓ → DLAT oligomerization↑ → mitochondrial dysfunction ● Metabolic: PI3K/AKT/mTOR/HIF-1α signaling↓ → Glycolysis↓ → OXPHOS↑	● Cell models: SMMC 7721 (human HCC), Hep G2 (human liver cancer), LO2 (normal liver) ● Animal models: Male BALB/c nude mice. Subcutaneous (for biodistribution), orthotopic (in situ HCC, injection into liver parenchyma) ● Administration: Intravenous injection, 4 mg/kg, on days 1, 6, 11 (3 doses) ● Radiation: 2 Gy (days 2, 7, 12)	● Relative tumor volume: 0.5-fold (vs. 94.3-fold control, in situ HCC, 2 Gy ● Tumor volume inhibition: significantly suppressed vs. other groups
CHP [295]	Cuproptosis/ ICD	● Cuproptosis: DLAT oligomerization; Cuproptosis inhibition (UK5099, 4T1, 80 µg/mL CHP): cell viability↑ from ≈ 50% to > 80% ● ICD: DAMPs (CRT↑, ATP↑, HMGB-1↑)	● Core: Copper and hafnium ions doped into a sodium tripolyphosphate (STPP) phosphate backbone ● Size: ≈20 nm ● Surface: Polyvinylpyrrolidone (PVP) coating ● Loading: None	● Physical: Hf-mediated high-Z radiosensitization ● Chemical: Cu2+ depletes GSH and amplifies ROS generation ● Biological: Induces cuproptosis; Reprograms the TME by alleviating hypoxia and repolarizing M2 to M1 macrophages, amplifying ICD	● Hf4+: DSB↑/γ-H2AX↑ ● Cu2+: GSH depletion (GSH↓) → ROS/•OH↑ ● Cuproptosis: FDX1↓/LIAS↓ → DLAT aggregation↑ → mitochondrial dysfunction	● Cell models: 4T1 (breast cancer), 3T3 (fibroblast), RAW264.7 (monocyte-macrophage), DC2.4 (dendritic cells) ● Animal models: Female BALB/c mice. Bilateral flank 4T1-tumor-bearing model ● Administration: Intravenous injection, 5 mg/kg ● Radiation: 2 Gy per fraction, 3 fractions	● Primary tumor growth inhibition rate: 82.1% (4T1 tumor, 2 Gy) ● Distant tumor growth inhibition rate: 77% (4T1 tumor, 2 Gy) ● Enhances immunogenicity: Mature DC↑; M1/M2↑
SPNCpG/Ca [305]	Ca-induced cell death/ ICD	● Ca-induced cell death: ROS generation↑. After SPNCpG/Ca treatment, the Ca2+ fluorescence signal was significantly enhanced within the cells. ΔΨm (JC-1) ↓ ● ICD: DAMPs (CRT↑, ATP↑, HMGB-1↑)	● Core: Semiconducting polymer (PFODBT) core ● Size: ≈62.7 nm ● Surface: DSPE-PEG-TK (thioketal linker) and CpG oligonucleotides. ● Loading: Ca2+ ions	● Physical: PFODBT works as radiosensitizer ● Chemical: SPN generates 1O2 under X-ray, enhancing RT and triggering CpG release; Ca2+ overloading. ● Biological: CpG activates antitumor immunity	● PFODBT: X-ray → 1O2 → CpG release; 1O2 also direct cytotoxicity ● Ca2+: mitochondrial permeability transition pore → ΔΨm↓ → ATP↓, ROS↑	● Cell models: 4T1 (breast cancer), BMSCs (bone marrow mesenchymal stem cells). ● Animal models: Female BALB/c mice. Bone metastasis model (4T1 cells injected into tibia of right hind limb). ● Administration: Intravenous injection (300 µg/mL, 200 µL). ● Radiation: 4 Gy, single dose, administered 12 h post-injection	● Tumor inhibition rate: 100% (bone metastatic tumor clearance, 4T1-based bone metastasis, 4 Gy) ● Mouse survival rate: 100% (4T1-based bone metastasis, 4 Gy, day 40). ● Lung metastatic nodules: 0 (SPNCpG/Ca + X-ray) vs. PBS 30.6 ± 2.1) ● Enhances immunogenicity: Mature DC cells↑; CD8+ T cells↑
ZnO–Au@mSiO2 [198]	Zinc-induced cell death/ ICD	● Zinc-induced cell death: Apoptosis rate↑ (ZnO-Au@mSiO2+X-ray 41.4% vs. PBS 3.6%, LLC cells); ROS↑; Lipid peroxidation ROS↑; Mitochondrial enlargement, swelling, and vacuolar degeneration; ΔΨm (JC-1) ↓ ● ICD: DAMPs (CRT↑, HMGB-1↑); cGAS-STING pathway activation leading to IFN-β production	● Core: Au@mSiO2. ● Size: ≈10–100 nm ● Surface: Polyacrylic acid coating ● Loading: DNAzyme (PD-L1-targeting) and ZnO (source of Zn2+)	● Physical: Au-mediated high-Z radiosensitization ● Chemical: X-ray induced ROS generation; Zn2+ release (pH-responsive) ● Biological: Zn2+-mediated DNAzyme activation for PD-L1 mRNA degradation; cGAS-STING pathway activation	● Au: X-ray → ROS → γ-H2AX↑, lipid peroxidation↑ ● Zn2+: activates DNAzyme → PD-L1 mRNA↓; activates cGAS-STING → p-TBK1↑, IFN-β↑; causes mitochondrial dysfunction (ΔΨm↓)	● Cell models: LLC (Lewis lung carcinoma) cells, LO2 (normal human liver) cells, DC 2.4 (dendritic) cells ● Animal models: Subcutaneous LLC tumor model in C57BL/6 mice ● Administration: Intravenous injection, 20 mg/kg, twice (days 0 and 7) ● Radiation: 4 Gy, fractionated (12 h post-injection, twice)	● Tumor growth inhibition: Complete inhibition (ZnO-Au@mSiO2+X-ray, LLC subcutaneous) ● Enhances immunogenicity: Mature DC cells↑ in lymph nodes and spleen; CD8+ T cells↑ in spleen
Mn3O4@GT [205]	Mn-induced cell death/ Pyroptosis	● Pyroptosis: Apoptosis rate (4T1, 6 Gy): 23.4% (vs. RT alone 17.6%); Elevated ROS and γ-H2AX expression; Cell swelling with large bubbles; GSDME-N↑; LDH↑, ATP, and IL-1β release	● Core: Mn3O4 nanoparticles ● Size: ≈200 nm. ● Surface: Tannic acid and GSDME protein ● Loading: Mn3O4 nanoparticles and GSDME protein	● Chemical: Mn3O4 depletes GSH and amplifies ROS ● Biological: GSDME hijacks caspase-3 to induce pyroptosis; Enhances anti-PD-L1 immunotherapy	● Mn3O4: GSH depletion → ROS↑ → γ-H2AX↑ ● RT: caspase-3 activation ● GSDME: cleaved by caspase-3 → GSDME-N → membrane pore formation → LDH, ATP, IL-1β release; COX-2↓, p-Akt↓ → PGE2↓, VEGF-A↓	● Cell models: 4T1 (mouse breast cancer), 4T1-luc (luciferase-expressing 4T1). ● Animal models: Female BALB/c mice. Subcutaneous unilateral 4T1 (for efficacy); bilateral 4T1 (primary right, distant left); co-injection model (irradiated 4T1 cells +4T1-luc cells) ● Administration: Intravenous injection, Mn3O4 0.2 mg/kg + GSDME 0.3 mg/kg, twice ● Radiation: 6 Gy (fractionated, 12 h post-injection, two cycles)	● Tumor regression: 92% complete regression (vs. RT alone, 4T1 subcutaneous, 6 Gy) ● Distant tumor regression: 85% (vs. IR alone, 4T1 bilateral model, 6 Gy, with PD-L1 blockade) ● Tumor repopulation (bioluminescence): 2.7-fold↓ (vs. IR+Mn3O4, 4T1-luc cells, 6 Gy) ● Enhances immunogenicity: Tregs↓ in distant tumor; CD8+ T cells↑ in distant tumor

AHSC: Atovaquone, Hafnium, Sabutoclax, and Chlorin e6; NPs: Nanoparticles; Bcl-2: B-cell lymphoma 2; Bcl-xL: B-cell lymphoma-extra large; Bak: BCL2 antagonist/killer; Bax: BCL2 associated X; ROS: Reactive oxygen species; DNA: Deoxyribonucleic acid; γ-H2AX: Phosphorylated histone H2AX; Hf: Hafnium; Ce6: Chlorin e6; PEG: Polyethylene glycol; Ato: Atovaquone; Sab: Sabutoclax; High Z: High atomic number; OCR: oxygen consumption rate; ICB: Immune checkpoint blockade; DSB: Double strand break; HNP: Hafnium-2-nitroimidazole@polyethylene glycol; ICD: Immunogenic cell death; CRT: Calreticulin; ATP: Adenosine triphosphate; HMGB1: High mobility group box 1; 2-nIm: 2-Nitroimidazole; DSPE-PEG5000: 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)−5000]; NO: Nitric oxide; RNS: Reactive nitrogen species; cGAMP: 2’,3’-Cyclic guanosine monophosphate-adenosine monophosphate; STING: Stimulator of interferon genes; TBK-1: Tank-binding kinase 1; IRF-3: Interferon regulatory factor 3; IFN-β: Interferon-β; CXCL10:C-X-C motif chemokine ligand 10; RT: Radiotherapy; s.c.: Subcutaneous; DCs: Dendritic cells; MOL: Metal-organic layer; SBU: Secondary building unit; Ir: Iridium; DBB: 4,4′-Di(4-benzoato)−2,2′-bipyridine; RDT: Radiodynamic therapy; TGI: Tumor growth inhibition indices; FA: Folic acid; ZIF-8: Zeolitic imidazolate framework-8; GSH: Glutathione; HIF-1α: Hypoxia-inducible factor 1 alpha; XRCC1: X-ray repair cross-complementing protein 1; ATM: Ataxia-telangiectasia mutated; ATR: Ataxia-telangiectasia and Rad3-related; Chk1: Checkpoint kinase 1; Chk2: Checkpoint kinase 2; p53: Tumor protein p53; caspase-3: Cysteine-aspartic acid protease 3; EdU: 5-ethynyl-2’-deoxyuridine; Bi₂S₃: Bismuth sulfide; 3BP: 3-Bromopyruvate; ATP: Adenosine triphosphate; HK-II: Hexokinase II; OXPHOS: Oxidative phosphorylation; LC3-II: Microtubule-associated protein 1 light chain 3; P62: Sequestosome-1; AgNPs: Silver nanoparticles; PVP: Polyvinylpyrrolidone; JNK/ERK: c-Jun N-terminal kinase/extracellular regulated protein kinases; ECFC: Endothelial colony forming cells; AuNRs: Gold nanorods; GPX4: Glutathione peroxidase 4; FTH: Ferritin heavy chain; NCOA4: Nuclear receptor coactivator 4; HO-1: Heme oxygenase 1; PSS: Poly(sodium 4-styrenesulfonate); HT: Hyperthermia; CYTO-ID: CYTO-ID^® Autophagy Detection Kit; N@VP: Verteporfin-loaded mPEG-PVGLIG-PCL nanoparticles; VP: Verteporfin; GSDMD: Gasdermin D; GSDMD-N: Gasdermin D N-terminal domain; IL-1β: Interleukin-1β; IL-18: Interleukin-18; mPEG-PVGLIG-PCL: Methoxypolyethyleneglycol-Pro-Val-Gly-Leu-Ile-Gly-polycaprolactone; NLRP3: NOD-like receptor thermal protein domain associated protein 3; IFN-γ: Interferon-gamma; IL-4: Interleukin-4; CDRT: Conventional-dose radiotherapy; HDRT: High-dose radiotherapy; Au: Gold; Ag: Silver; Bi: Bismuth; AgBiS2: Silver bismuth sulfide; GSDME: Gasdermin E; PBS: Phosphate-buffered saline; DAMP: Damage-associated molecular pattern; PEG-C18PMH: Polyethylene glycol-grafted poly(maleic anhydride-alt-1-octadecene); 53BP1: p53-binding protein 1; DAC: decitabine; O-HONs: oleate-PEG modified hafnium oxide nanoparticles; GSDME: gasdermin E; LDH: lactate dehydrogenase; HfO₂: hafnium dioxide; OA-PEG: oleate-polyethylene glycol; PWE: polyphenol-based metal-phenolic nanocoordinator containing tungsten and epigallocatechin-3-gallate; EGCG: epigallocatechin-3-gallate; W⁶⁺: tungsten(VI) ion; PEGylated: polyethylene glycolylated; mPEG-b-Pho: methoxypolyethylene glycol-block-polyphenol; DNMT: DNA methyltransferase; Phy: Physcion; PLGdH: PEGylated layered gadolinium hydroxide nanosheets; CHOP: C/EBP homologous protein; P-STING: Phosphorylated STING; P-IRF3: Phosphorylated interferon regulatory factor 3; LGdH: Layered gadolinium hydroxides; PPP: Pentose phosphate pathway; NADPH: Nicotinamide adenine dinucleotide phosphate; Gd: Gadolinium; 6PGD: 6-Phosphogluconate dehydrogenase; αPD-L1: Anti-programmed cell death ligand 1 antibody; CS-Mn: Chitosan-manganese coordinated microparticles; P-TBK1: Phosphorylated TANK-binding kinase 1; IL-15: Interleukin-15; TNF-α: Tumor necrosis factor alpha; Th-Ir-DBB/Dig: Thorium-iridium-dibenzoato-bipyridine/digitonin nanosystem; GSSG/GSH: Glutathione disulfide/reduced glutathione; BAY-876: BAY-876 (GLUT1 inhibitor); F-actin: Filamentous actin; Th₆O₄(OH)₄: Thorium hexanuclear oxo hydroxide cluster; SBUs: Secondary building units; Ir-DBB: Iridium complex with 4,4′-di(4-benzoato)−2,2′-bipyridine ligand; GLUT1: Glucose transporter 1; IL-6: Interleukin-6; GOx: Glucose oxidase; FBSP: Fe-Bi disulfide-bridged metal-organic framework; BODIPY C11: Boron-dipyrromethene C11 (lipid peroxidation sensor); PTGS2: Prostaglandin-endoperoxide synthase 2 (also known as COX2); Fe@Pdots-siRNA: Iron-doped semiconducting polymer dots loaded with small interfering RNA; SLC7A11: Solute carrier family 7 member 11 (also known as xCT)

MEH-PPV: Poly[2-methoxy-5-(2-ethylhexyloxy)−1,4-phenylenevinylene]; NPC: Nasopharyngeal carcinoma; CNE2: Human nasopharyngeal carcinoma cell line; siRNA: Small interfering RNA; circADARB1: Circular RNA ADARB1; HSP90B1: Heat shock protein 90 beta family member 1; Fe-doped: Iron-doped; 4-HNE: 4-Hydroxynonenal; dp-PEG-mal: Diphosphate-polyethylene glycol-maleimide; SH-HA: Sulfhydryl-hyaluronic acid; IR: Ionizing radiation; PFCN: PEGylated Fe₃O₄-decorated CaWO₄ nanoparticles; MDA: Malondialdehyde; CaWO₄: Calcium tungstate; LPO: Lipid peroxidation; HA: Hyaluronic acid; DLAT: Dihydrolipoamide S-acetyltransferase; FDX1: Ferredoxin 1; LIAS: Lipoyl synthase; PVP: Polyvinylpyrrolidone; PEG-SS-PEG: Polyethylene glycol-disulfide-polyethylene glycol; 2-nIm: 2-Nitroimidazole; NO: Nitric oxide; OXPHOS: Oxidative phosphorylation; PI3K: Phosphoinositide 3-kinase; AKT: Protein kinase B; mTOR: Mammalian target of rapamycin; CHP: Cu/Hf-doped phosphate nanostimulator; STPP: Sodium tripolyphosphate; TME: Tumor microenvironment; M2: M2 macrophage; M1: M1 macrophage; SPNCpG/Ca: Semiconducting polymer nanoparticle loaded with CpG oligonucleotides and calcium ions; PFODBT: Poly[(9,9-dioctylfluorenyl-2,7-diyl)-co-(4,7-di-2-thienyl-2,1,3-benzothiadiazole)]; DSPE-PEG-TK: 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-polyethylene glycol-thioketal; CpG: Cytosine-phosphate-guanine oligonucleotide; SPN: Semiconducting polymer nanoparticle; ZnO: Zinc oxide; Au: Gold; mSiO₂: Mesoporous silica; Mn₃O₄@GT: Manganese tetroxide@Gasdermin E-tannic acid nanoassembly; COX-2: Cyclooxygenase-2; P-Akt: Phosphorylated protein kinase B; PGE₂: Prostaglandin E₂; VEGF-A: Vascular endothelial growth factor A

Apoptosis

Apoptosis is the most well established cell death mode and a primary mechanism for radiation-induced tumor eradication. Executed via intrinsic (mitochondrial) and extrinsic (death receptor) pathways, apoptosis is defined by chromatin condensation and apoptotic body formation [227]. The extrinsic route is initiated by ligands such as Fas binding to death receptors to activate Caspase-8 and Caspase-3. Conversely, RT predominantly triggers the intrinsic pathway through DNA DSBs, leading to mitochondrial membrane potential collapse, cytochrome c release, and the subsequent activation of Caspase-9 and Caspase-3 (Fig. 4A).

Fig. 4.

Nano-radiosensitization via apoptosis, autophagy, pyroptosis, and disulfidptosis. (A) Apoptosis: Radiosensitization triggers both extrinsic and intrinsic apoptotic pathways. NMs and radiation activate death receptors (FAS) and induce mitochondrial outer membrane permeabilization, leading to the release of cytochrome c and activation of the Caspase-9/3 cascade. Crosstalk is mediated by Casp8-induced tBid formation. Cytosolic leakage of dsDNA activates the cGAS-STING-IRF3 axis, promoting Type I Interferon secretion. (B) Autophagy: Supraphysiological ROS levels generated by NMs and radiation activate ERK/JNK signaling, leading to mTOR inhibition and ULK1 complex activation. This initiates autophagosome formation, characterized by LC3-I to LC3-II conversion and ATG5/12/16 recruitment. NMs can either induce damaging autophagy or block protective autophagic flux (via inhibitors like 3-MA or TAT-GNSs) to enhance cytotoxicity. (C) Pyroptosis: ROS-mediated activation of the NLRP3 inflammasome triggers the canonical Casp1-GSDMD pathway, releasing pro-inflammatory cytokines (IL-18, IL-1β). Simultaneously, NMs delivering DNMT inhibitors (PWE) reverse GSDME silencing via demethylation; subsequently, radiation-activated Casp3 cleaves GSDME into N-GSDME pores, converting apoptosis into pyroptosis and releasing DAMPs to activate DC and T-cell mediated immunity. (D) Disulfidptosis: NMs interfere with glucose uptake (GLUT1 inhibition) or metabolism, leading to NADPH depletion. Under SLC7A11-mediated cystine overload, the lack of reducing power (NADPH) results in disulfide stress and abnormal disulfide bond formation among cytoskeletal proteins, culminating in F-actin network collapse and metabolic cell death

Nanomaterials enhance this process through multifaceted strategies. High-Z elements (such as Au [192], Gd [228], Pt [229], and Hf [219]) can intensify energy deposition to exacerbate terminal DNA lesions. For example, Hf-nIm@PEG nanoparticles utilize Hf⁴⁺ for physical sensitization and 2-nitroimidazole (2-nIm) to release NO under irradiation. This synergy generates reactive nitrogen species and activates the cGAS-STING signaling to facilitate immune-mediated apoptosis [96]. Precision platforms also activate the extrinsic pathway by delivering death ligands or agonists; SeC@MSNs-Tf/TAT facilitates Fas/FasL signaling [230], while SeD@MSNs-FA triggers ROS-dependent apoptosis and S-phase arrest [166]. Furthermore, Pt@Bi₂Se₃-RGD (PBR) assemblies target the amphiregulin (Areg)/EGFR/Bcl-2 axis to remodel radiosensitivity [231].

Mitochondrial regulation, governed by the Bcl-2 family, remains a primary target for nano-intervention. FA-Mn₃O₄@ZIF-8 nanoparticles catalyze H₂O₂ decomposition to generate ROS, which induces mitochondrial membrane potential (ΔΨm) collapse and cytochrome c release [201]. Other innovative designs include 1P-SmacN7 supramolecular antagonists that neutralize inhibitor of apoptosis proteins to activate the caspase cascade [232]. Additionally, AHSC NPs provide a multi-pronged attack by combining Hf-enhanced radiation, chlorin e6 (Ce6)-mediated ROS generation, atovaquone (Ato) induced respiratory chain inhibition, and Sabutoclax (Sab) mediated Bcl-2 suppression to consolidate apoptotic signaling [233].

In contrast to established apoptosis, PANoptosis is an emerging inflammatory cell death mode characterized by the assembly of the PANoptosome complex. This molecular scaffold integrates signaling pathways from pyroptosis, apoptosis, and necroptosis by involving the caspase family, RIPK1/3, and MLKL to achieve synergistic activation of multiple death mechanisms [234–238]. Nanomaterials have been reported to trigger PANoptosis through several strategies, such as enhanced drug delivery [239], stimulus-triggered ROS generation [240–244], enzyme-mimetic activities [245, 246], and the disruption of ionic homeostasis [247].

Notably, emerging evidence suggests that nano-radiosensitizers may act as potential inducers of PANoptosis. Zhang et al. (2025) developed a diselenide-bonded COF, PorSe-CuPt@CBL, which provides both physical and biological radiosensitization [248]. By remodeling the immunosuppressive tumor microenvironment, this platform was reported to trigger PANoptosis and potentially amplify RT efficacy. Although PANoptosis, if confirmed, could serve as a “multi-hit” mechanism capable of dismantling tumor defenses, its intersection with radiosensitization remains largely unexplored. The existence of only one initial study to date highlights a vast frontier for future investigation. Moreover, as an emerging and still poorly understood concept, its practical application in clinical radiosensitization remains largely exploratory, and the lack of specific pharmacological inhibitors or genetic tools to dissect PANoptosis from other death pathways poses a significant validation challenge at this stage.

In contrast, necroptosis is a regulated programmed cell death modality that serves as an alternative pathway when apoptosis is impaired, such as in the absence of Caspase-8 [249]. This process is primarily executed by RIPK1, RIPK3, and the downstream effector MLKL [250–252]. While various sonodynamic, photodynamic, and photothermal nanosystems have been developed for oncological applications [253–257], only a limited number of nanomaterials have been reported to specifically induce tumor cell necroptosis through a radiosensitization-based strategy.

Autophagy

Autophagy is a highly conserved cellular self-degradation process where lysosomes eliminate damaged organelles and misfolded proteins to maintain intracellular homeostasis [258]. This mechanism is strictly regulated by complexes including unc-51-like kinase 1 (ULK1) and class III phosphatidylinositol 3-kinase, involving the lipidation of microtubule-associated protein 1 light chain 3 (LC3) to form autophagosomes for resource recycling [259]. While tumor cells often activate protective autophagy to survive radiation-induced DNA damage and ROS, excessive or persistent activation beyond a tolerance threshold can trigger autophagic cell death or amplify cytotoxicity through cross-talk with other death pathways.

Nanomaterials can be engineered to induce damaging autophagy, typically driven by supraphysiological ROS levels [260] (Fig. 4B). For instance, silver nanoparticles (AgNPs) promote LC3-I to LC3-II conversion and p62 degradation by activating extracellular signal-regulated kinase (ERK) and JNK signaling [261]. Similarly, Bi₂S₃−3BP nanoparticles and other ROS-generating platforms, such as FePt/GO [260], AuP NPs [262], AgNPs [263], CuO NPs [264], have been shown to elevate the LC3-II/LC3-I ratio to exacerbate RT-induced damage [121].

Inhibition of protective autophagy is another important strategy for reversing radioresistance. Nanomaterials can directly block autophagic flux; ECFC-AuNRs downregulate Beclin-1 and trigger p62 accumulation in breast cancer models [115], while TAT-GNSs hinder autophagosome degradation to improve RT outcomes [265]. Furthermore, nanocarriers can precisely deliver inhibitors to the tumor site. For example, Ag@MSNs loaded with 3-methyladenine (3-MA) effectively inhibit autophagy to amplify oxidative stress in glioma cells, thereby achieving radiosensitization in vitro and in vivo [266]. A limitation of current autophagy-targeting strategies is that the threshold distinguishing protective from detrimental autophagy remains poorly defined.

Pyroptosis

Pyroptosis is a programmed necrotic cell death mediated by the Gasdermin family, characterized by cellular swelling, membrane rupture, and the release of pro-inflammatory cytokines such as interleukin-1β (IL-1β) and interleukin-18 (IL-18) [267]. The core mechanism involves Gasdermin activation via canonical or non-canonical inflammasome pathways [268]. Nanomaterials trigger this process by regulating Gasdermin E (GSDME) and promoting the infiltration of DCs, macrophages, type 1 helper T (Th1) cells, and CD8⁺ T cells (Fig. 4C).

Nanomaterials amplify radiation-induced pyroptosis by serving as catalysts for ROS generation. For instance, N@VP nanoparticles developed by Zhang et al. have been reported to generate large amounts of ROS, particularly ¹O₂, upon X-ray excitation to trigger Gasdermin D (GSDMD) mediated pyroptosis [222]. Similarly, nanomaterials such as Au@AgBiS₂-PEG [269] and Au-Pt@Mxene@PVA [270] facilitate the formation of pore-forming GSDME N-terminal fragments (GSDME-N), leading to the release of HMGB1, ATP, and CRT to initiate ICD.

Additionally, nanocarriers can deliver epigenetic modulators to reverse pyroptotic silencing. The PWE platform delivers epigallocatechin gallate to counteract GSDME hypermethylation [271]. Likewise, the DAC@O-HONs system carries decitabine to inhibit DNA methyltransferases, thereby reversing GSDME silencing in triple-negative breast cancer and effectively converting radiation-induced apoptosis into pro-inflammatory pyroptosis [203]. Despite these promising results, the translational potential of pyroptosis-inducing nanosystems is limited by the fact that GSDME expression is often silenced in many cancer types, and the systemic induction of pyroptosis may cause off-target inflammatory toxicity.

Disulfidptosis

Disulfidptosis, first described in 2023, is a novel cell death mode triggered by excessive intracellular cystine accumulation and subsequent disulfide stress [272]. This process is characterized by catastrophic disulfide crosslinking of cytoskeletal proteins, primarily the actin network, leading to morphological collapse. Typically occurring under glucose deprivation and high SLC7A11 expression, glucose starvation inhibits the pentose phosphate pathway, depleting NADPH and preventing cystine reduction. The resulting accumulation of disulfides induces aberrant disulfide bond formation in the actin cytoskeleton, culminating in structural disintegration [273].

Nanomaterials have been explored to accelerate disulfidptosis by delivering glucose-depleting agents or impeding glucose uptake to induce NADPH deficiency (Fig. 4D). For example, digitonin-loaded Th-Ir-DBB/Dig disrupts membrane integrity and inhibits glucose transporter 1, triggering cystine accumulation and actin disintegration [199]. Preliminary evidence suggests that disulfidptosis sensitizes cells to ROS and upregulates surface CRT, which enhances phagocytosis by antigen-presenting cells (APCs) to activate T cells. Furthermore, GOx-modified FBSP MOFs deplete glucose to generate disulfide stress and initiate ICD, showing potential to potentiate anti-tumor immunity and RT efficacy [175]. However, disulfidptosis research is still in its infancy, and only a handful of nanomaterial-based studies exist.

Metal ion-related cell death

Intracellular ion homeostasis is indispensable for maintaining cellular metabolism and survival. Tumor cells adapt to rapid proliferation by reprogramming ion transport systems, which simultaneously creates a unique vulnerability to ionic fluctuations [274]. Metal ions, as essential trace elements, regulate critical physiological processes, including signal transduction and enzyme catalysis [275]. Consequently, nanomedicine-mediated disruption of this balance has birthed “ion interference therapy”, a novel oncology paradigm that triggers potent stress responses and cell death [184, 274]. This approach gained increased attention following the 2022 discovery of cuproptosis, which catalyzed the integration of metal ion regulation with RT [276]. Metal-based nanomaterials exhibit distinct advantages in targeted therapy by leveraging the EPR effect and surface functionalization [274]. Crucially, these platforms can function as dual-action agents: they serve as efficient radiosensitizers to enhance energy deposition while concurrently releasing metal ions to induce specific death pathways, potentially achieving robust synergistic sensitization. One major limitation of ion interference therapy is the potential for systemic metal toxicity, as the biodistribution and long-term clearance of metal-based nanomaterials are often not comprehensively evaluated in preclinical studies.

Ferroptosis

Ferroptosis is an iron-dependent cell death characterized by LPO, mitochondrial dysfunction, and GPX4 inactivation [277]. The biochemical hallmark involves iron-catalyzed Fenton reactions that generate •OH radicals, which attack PUFAs to initiate a chain reaction of LPO [278]. This process is exacerbated by GSH depletion and the accumulation of toxic byproducts, such as 4-hydroxy-2-nonenal (4-HNE) and malondialdehyde (MDA), which activate the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway. Cancer cells exhibit a high dependence on iron and glutamine metabolism, rendering them uniquely susceptible to ferroptosis-inducing strategies. Beyond direct cytotoxicity, ferroptosis promotes dendritic cell maturation and T cell infiltration, effectively enhancing antitumor immunity.

RT has been suggested to initiate a positive feedback loop by increasing ROS levels and upregulating Acyl-CoA Synthetase Long Chain Family Member 4 (ACSL4), a prerequisite for ferroptosis [82]. Nanomaterials further amplify this synergy. Several iron-based nanomaterials (such as Fe₃O₄ nanoparticles [147], FeNDs nanoparticles [279], Fe@Pdots-siRNA [280], BZAMH [176]) directly elevate intracellular iron to catalyze H₂O₂ into ROS, thereby inhibiting the SLC7A11-GSH-GPX4 axis and promoting ferroptosis (Fig. 5A).

Fig. 5.

Nano-radiosensitization via ferroptosis, cuproptosis, and immunogenic cell death. (A) Induction of ferroptosis: RT upregulates the expression of ACSL4, promoting the incorporation of PUFAs into membrane phospholipids. Concurrently, NMs enhance cellular iron uptake and catalyze the Fenton reaction, converting H₂O₂ into highly reactive •OH. NMs further sensitize cells by inhibiting key antioxidant defense axes, including the solute carrier SLC7A11/GSH/GPX4 pathway and the FSP1 system. The resultant accumulation of lipid peroxidation products (e.g., 4-HNE, MDA) leads to irreversible membrane rupture and the release of immunostimulatory signals that activate DCs and T cells. (B) Induction of cuproptosis: RT facilitates copper influx by upregulating the copper transporter SLC31A1 via the transcription factor ETS1. Within mitochondria, redox perturbations induced by both RT and NMs promote the FDX1-mediated reduction of Cu²⁺ to Cu⁺. The toxic Cu⁺ ions bind to lipoylated components of the TCA cycle (e.g., dihydrolipoamide S-acetyltransferase), triggering their aberrant oligomerization and the subsequent loss of iron-sulfur (Fe-S) cluster proteins. This cascade results in acute mitochondrial metabolic collapse, loss of membrane integrity, and cell death. (C) Induction of ICD: The combined action of RT and NMs elicits a potent immunogenic cascade through the emission of DAMPs: (i) ER stress induces the translocation and surface CRT. (ii) Mitochondrial damage promotes the release of ATP via pannexin-1 channels. (iii) Nuclear damage enables the passive secretion of HMGB1. (D) Cytosolic dsDNA activates the cGAS–STING pathway, leading to the phosphorylation of IRF3 and the production of IFN-I. Collectively, these signals drive the maturation of DCs, enhance tumor infiltration by cytotoxic T cells, and promote the repolarization of immunosuppressive M2-type tumor-associated macrophages toward the anti-tumor M1 phenotype.

To mitigate potential systemic toxicity from exogenous iron, recent designs focus on mobilizing endogenous iron pools [281, 282]. For instance, ovs-MnO₂ induces the release of intracellular Fe²⁺ to trigger LPO [117]. Similarly, the nanomaterial iCoDMSN can disrupt Kelch-like ECH-associated protein 1 (KEAP1) to activate Nrf2 and upregulate the expression of Heme Oxygenase-1, releasing iron from heme degradation [283]. Furthermore, nanocarriers can load specific inducers to promote ferroptosis. Hollow Bi₂Se₃ nanomaterials deliver RSL3 to upregulate ferroptosis markers including ACSL4 and prostaglandin-endoperoxide synthase 2 [177]. Gao et al. integrated the GSH-depleting sensitizer 808-NM2 into 8HSA8 nanoparticles, promoting iron accumulation and inhibiting the SLC7A11-GSH-GPX4 axis to improve RT outcomes [202].

Non-iron-dependent pathways also facilitate ferroptosis through ROS and metabolism disruption. Au/CuNDs utilize Cu⁺ to promote ROS generation and GSH oxidation, indirectly fueling LPO [90]. CoO@BSA nanoparticles initiate Fenton-like reactions and interfere with the KEAP1/NRF2/HMOX1 pathway by reducing KEAP1 expression while upregulating NRF2 and HMOX1 to enhance iron accumulation [283]. Similarly, GPGP nanoplatforms utilize catalytic POD and CAT-like activities to trigger LPO [140]. BMBs nanomaterials mimic POD, GPX, and NADPH dehydrogenase (NDH) activities to deplete GSH and NADPH in the acidic TME, enabling the simultaneous inhibition of GPX4 and ferroptosis suppressor protein 1 (FSP1) [136].

TME-responsive iron-based nanomaterials further ensure localized release and minimized systemic reactions. Fe-doped Mn₃O₄ nanoparticles degrade in the acidic TME to release Fe²⁺, catalyzing LPO and generating highly cytotoxic 4-HNE [89]. These particles concurrently deplete GSH and inhibit GPX4 to further promote LPO. Additionally, Fe₃O₄-decorated CaWO₄ nanoparticles utilize X-ray-induced ultraviolet light to promote the reduction of Fe³⁺ to Fe²⁺, and subsequently trigger a cascade Fenton reaction that generates toxic •OH to induce tumor cell ferroptosis [133].

A critical limitation of ferroptosis-focused radiosensitization is the reliance on in vitro LPO or GPX4 activity assays, whereas in vivo evidence of ferroptosis as the dominant death mechanism remains scarce. Moreover, the potential for off-target iron accumulation in the heart, liver, and spleen raises safety concerns that have not been systematically addressed.

Cuproptosis

Cuproptosis is a copper-dependent death mode characterized by mitochondrial metabolic disturbances, specifically the abnormal oligomerization of lipoylated proteins in the TCA cycle and the loss of iron sulfur (Fe-S) cluster proteins [276]. Intracellular copper homeostasis is governed by a sophisticated network, including copper transporter 1 and 2 (CTR1 and CTR2, also known as SLC31A1 and SLC31A2) for uptake, metallothioneins and GSH for sequestering, and P-type ATPases (ATP7A and ATP7B) for efflux [284]. Unlike apoptosis or ferroptosis, cuproptosis results from a direct metabolic crisis rather than caspase activation or LPO.

The mechanisms of cuproptosis involve both mitochondrial and cytoplasmic pathways. In the mitochondrial route, FDX1 reduces Cu²⁺ to Cu⁺, which binds lipoylated DLAT to induce lethal oligomerization and destabilize iron sulfur (Fe-S) clusters, causing metabolic collapse [276, 285]. Conversely, the mitochondrial independent pathway utilizes disulfiram copper complexes to aggregate valosin containing protein (VCP, or p97) and its cofactor Npl4. This process inhibits the ubiquitin proteasome system, triggering severe cytoplasmic proteotoxic stress independent of mitochondrial metabolism [286–290] (Fig. 5B).

Notably, tumor cells frequently undergo copper metabolism reprogramming, characterized by the upregulation of CTR1 and downregulation of copper efflux pumps, specifically ATPase copper transporting alpha and beta (ATP7A/B), resulting in elevated intracellular copper levels [291, 292]. This metabolic vulnerability makes cuproptosis a promising therapeutic target. RT has been reported to further potentiate cuproptosis by upregulating CTR1 through E26 transformation-specific sequence 1 (ETS1) and depleting mitochondrial GSH, which exacerbates copper toxicity [293]. Nanotechnology offers distinct advantages for inducing cuproptosis by serving as targeted delivery systems that enhance copper bioavailability and cellular uptake. Various platforms, including MOFs [294], copper-based nanoparticles [200], and composite nanomaterials [295], facilitate efficient intracellular copper release. While Cu⁺ directly binds DLAT and FDX1 to trigger cell death [101], and Cu²⁺ enhances the RT response by inducing ROS and further depleting GSH [295]. Crucially, cuproptosis has been shown to initiate ICD by releasing tumor-associated antigens, thereby activating a systemic antitumor immune response [101, 295].

Recent strategies prioritize optimizing the biocompatibility and targeting of copper nanoplatforms to minimize systemic toxicity. For example, PWCu nanocapsules utilize radiation-induced water radiolysis to reduce Cu²⁺ to Cu⁺, which binds lipoylated proteins to trigger lethal aggregation and disrupt iron sulfur (Fe-S) clusters in the TCA cycle, thus inducing cuproptosis [178]. Additionally, platelet membrane-coated PDA@Cu nanoparticles leverage biomimetic camouflage for tumor targeting and exploit the acidic TME for localized Cu²⁺ release [200]. Intracellular reduction by GSH then converts Cu²⁺ to Cu⁺ to initiate cuproptosis and catalyze ROS generation, potentially potentiating the RT response. Despite these advances, the clinical translation of cuproptosis-based radiosensitization is hindered by the lack of specific non-invasive biomarkers to monitor cuproptosis in vivo and the fact that most copper-based nanomaterials exhibit poor stability in physiological fluids, leading to premature copper release.

Other metal ions-related tumor cell death

Calcium overload

Calcium is the most abundant metal in the human body, with free Ca²⁺ serving as a universal second messenger that regulates signal transduction and cellular metabolism [296]. While the ER, mitochondria, and sarcoplasmic reticulum maintain ionic balance, pathological calcium overload occurs when intracellular concentrations surpass regulatory thresholds, triggering mitochondrial collapse and ER stress [297]. Mechanistically, calcium overload disrupts the mitochondrial permeability transition (PT) pore, leading to ΔΨm collapse and cytochrome c release to activate the apoptotic cascade [298, 299]. Concurrently, excessive Ca²⁺ inhibits electron transport chain complexes to amplify oxidative stress through O₂•⁻ generation and phospholipase activation [300]. Beyond direct cytotoxicity, calcium overload initiates immunomodulatory effects by inducing CRT exposure and the release of DAMPs. These signals enhance the antigen-presenting capability of DCs and polarize tumor associated macrophages toward the antitumor M1 phenotype, thereby augmenting immunotherapy efficacy [301, 302]. By integrating these multifaceted death pathways, nanomedicine-mediated calcium overload may synergize with RT-induced DNA damage through strategic exogenous delivery or endogenous release activation.

Nanocarriers facilitate the direct delivery of exogenous calcium or calcium compounds to elevate intracellular Ca²⁺ levels. For example, hafnium-doped hydroxyapatite nanoparticles undergo acidic dissolution to release Ca²⁺ [303], while CFM nanocomposites utilize calcium peroxide (CaO₂) decomposes to inflict mitochondrial damage and enhance RT efficacy [304]. Additionally, SPNCpG/Ca platforms directly deliver Ca²⁺ to impair mitochondrial function and trigger ER stress, whereas the co-delivered cytosine-phosphate-guanine (CpG) oligonucleotides activate DCs to prime antitumor immunity [305].

Alternatively, advanced nanomaterials modulate endogenous calcium release from intracellular stores. Unlike exogenous delivery, the dual-channel nanoregulator CAP-P-NO, integrating capsaicin and NO, orchestrates a dual-source calcium overload [306]. This system activates the transient receptor potential vanilloid 1 channel for extracellular influx and sensitizes ryanodine receptors via S-nitrosylation for ER release. The resulting calcium disequilibrium triggers a ROS surge and suppresses DNA repair, as evidenced by elevated γ-H2AX expression, thereby synergistically amplifying radiation-induced DNA damage. A key limitation of calcium overload strategies is that the threshold for apoptosis versus necrosis is narrow, and excessive calcium release can cause acute toxicity. Moreover, the role of calcium signaling in the tumor microenvironment is context-dependent, and some tumors may adapt to calcium stress via upregulated efflux pumps.

Mn-induced tumor cell death

Mn, a versatile transition element, regulates critical physiological functions ranging from enzyme activity to oxidative stress [307–311]. Among its oxidation states, divalent manganese (Mn²⁺) is the primary bioactive species, utilizing its redox activity and coordination ability to serve as a vital catalytic cofactor [312]. Beyond its utility as a contrast agent for multispectral optoacoustic tomography, Mn-based nanomaterials enhance radiosensitivity through multifaceted mechanisms [117, 313, 314]. These platforms are thought to amplify RT efficacy through multiple synergistic mechanisms.

Mn-based platforms also remodel the tumor immune microenvironment by activating the cGAS-STING axis [311]. During RT, double-stranded DNA released from tumor cells is recognized by cGAS, an interaction markedly enhanced by Mn²⁺ binding [89, 195, 315]. This efficiently catalyzes the synthesis of the second messenger 2’3’-cyclic GMP-AMP (cGAMP), triggering a signaling cascade that promotes DC maturation and the secretion of IFN-I, interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α) [316]. Consequently, this process activates innate and adaptive immunity while inducing ICD and pyroptosis [205]. For instance, CS-Mn microparticles capture DNA with 94.2% efficiency to synergistically activate cGAS-STING and enhance RT-induced immunity [195]. As natural adjuvants, Mn-based materials are frequently integrated with high-Z elements like Hf and Bi (HfMnH and CD_{Bi, Mn−FA}@CaP) to realize synchronous sensitization and sustained STING activation [104, 196]. Furthermore, Mn²⁺ inhibits the immunosuppressive functions of MDSCs [89]. Nanoparticles like Fe-doped Mn₃O₄ and UiO@MnS reduce M2 macrophage polarization to facilitate CD8⁺ T cell infiltration and DC maturation, transforming the TME from an immunosuppressive to an immuno-activated state [89, 119].

Mn-containing nanoparticles also exhibit significant POD, GPX, and NDH-like activities [104, 189, 201, 317]. Systems such as FA-Mn₃O₄@ZIF-8 catalyze Fenton-like reactions to generate toxic •OH, thereby increasing oxidative stress and cell cycle arrest [119, 201]. Additionally, Mn₃O₄ reacts with GSH to produce Mn²⁺ and GSSG, depleting antioxidant defenses to exacerbate RT-induced ROS accumulation [201, 205, 317]. MnO_x can further mimic multifaceted enzyme activities to deplete intracellular GSH and NADPH, inhibiting GPX4 and FSP1 to induce ferroptosis [136]. Although Mn-based nanomaterials show promise, their clinical application is complicated by the potential neurotoxicity of excess Mn²⁺, as chronic exposure can lead to manganism. Furthermore, the optimal dosage for achieving radiosensitization without systemic adverse effects has not been established.

Zn-induced tumor cell death

Zn is an essential trace element vital for protein synthesis, DNA repair, and immune signaling [318, 319]. While tumor cells sequester Zn to support proliferation, supraphysiological levels trigger terminal death signals and apoptosis [318, 320].

In RT, Zn may potentiate therapeutic efficacy by modulating immune responses, enhancing oxidative stress, and inducing mitochondrial dysfunction [321]. Specifically, ZnO–Au@mSiO₂ nanocomposite release Zn²⁺ serves as a dual activator of DNAzyme and cGAS-STING thereby suppressing PD-L1 expression and priming antitumor immunity [198]. Moreover, zinc oxide nanoparticles (ZnO NPs) generate electron-hole pairs upon X-ray exposure to elevate ROS levels [198, 322]. Zn also dictates cell cycle dynamics, as seen with Zn@Alg-Dox NPs induce G2/M phase arrest to intensify the radiation response [323]. Despite these preliminary advantages, zinc-mediated radiosensitization remains relatively under-explored, and the mechanisms of Zn-induced cell death beyond apoptosis are poorly characterized.

ICD

ICD is a form of cell death that elicits an adaptive immune response by transforming dying tumor cells into “in situ vaccines” [324]. It can be triggered by diverse cell death modes, including immunogenic apoptosis, necroptosis, ferroptosis, and cuproptosis [325]. The hallmarks of ICD involve the spatiotemporal exposure or release of DAMPs, such as cell-surface calreticulin, extracellular ATP, and HMGB1, alongside heat shock proteins and IFNs. These signals orchestrate an immune cascade where DCs recognize DAMPs through pattern recognition receptors, undergo maturation, and present tumor-derived antigens to T cells within lymph nodes. This process recruits cytotoxic T lymphocytes and Th1 cells to the tumor site, establishing systemic antitumor immunity and long-term surveillance to combat metastasis and recurrence.

Conventional RT effectively induces ICD by generating ROS and DNA damage, which trigger ER stress and mitochondrial dysfunction to facilitate DAMPs translocation and the subsequent abscopal effect [326]. Nevertheless, the clinical efficacy of radiation-induced ICD is frequently hindered by the immunosuppressive TME. Factors such as Tregs, MDSCs, and upregulated PD-L1 can neutralize DAMPs-mediated activation, leading to an attenuated immune response.

Nanomaterials circumvent these limitations by enhancing ICD induction across multiple dimensions. They may potentiate RT by amplifying DAMPs release, directly regulating biomarker exposure, and activating the cGAS-STING pathway through modulated DNA repair. Furthermore, nanotechnology reconfigures the TME to synergize with ICD-induced responses, thereby extending the therapeutic breadth and depth of traditional RT (Fig. 5C).

ICD is centered on the release of DAMPs as primary signals to initiate adaptive immunity. Given that conventional RT often fails to elicit robust ICD signals, nanomaterials leverage physical and chemical sensitization to amplify cellular damage and the exposure of DAMPs, such as nuclear HMGB1 and cytoplasmic heat shock proteins (HSPs). For example, high-Z element (gadolinium oxide) in Phy@PLGdH nanosheets intensifies radiation deposition to elevate ROS and DNA DSBs. The resulting fragments activate cGAS-STING to secrete interferon-beta (IFN-β), recruiting APCs into the TME for antigen cross-presentation and CD8⁺ T cell activation [97]. Furthermore, the cuproptosis-inducing nanocapsule PWCu releases Cu⁺ upon irradiation to cause mitochondrial dysfunction and lethal protein lipoylation, facilitating the extracellular translocation of CRT and HMGB1 to prime systemic antitumor immunity [178].

Beyond indirect sensitization, nanomaterials directly modulate core ICD-associated molecules such as CRT, HMGB1, ATP, and HSPs. By precisely regulating the expression and secretion of these signals, nanoplatforms significantly amplify ICD induction. For instance, AuNPs combined with RT upregulate CRT, HMGB1, and ATP in glioblastoma cells to enhance therapeutic efficacy [193]. Furthermore, AmGd-NPs loaded with cluster of differentiation 73 (CD73) inhibitors suppress enzymatic activity to prevent extracellular ATP degradation. The resulting ATP accumulation facilitates DC activation and antigen presentation, ultimately priming CD⁸⁺ T cell-mediated antitumor immunity [228].

Cytoplasmic dsDNA is the pivotal signal for activating the cGAS-STING pathway to initiate immune responses. Nanomaterials potentiate this axis by both amplifying RT-induced DNA damage to increase dsDNA levels and acting as direct agonists. Specifically, Hf-nIm@PEG (HNP) nanoparticles utilize NO to inhibit DNA repair while Hf⁴⁺ activates the cGAS-STING pathway, triggering the maturation of bone marrow-derived dendritic cells (BMDCs) and the secretion of cytokines such as C-X-C motif chemokine ligand 10 (CXCL10) and IFN-β [96]. Additionally, manganese-based nanomaterials (e.g., Mn₃O₄ nanoparticles [89], CS-Mn microparticles [205] and zinc-based nanoparticles [198, 322]) have shown the ability to directly activate the STING pathway.

Nanocarriers also ensure the precise delivery of agonists to amplify the immune-sensitizing effects of RT. For instance, the two-dimensional metal-organic layer platform, cGAMP/MOL, facilitates the phosphate-triggered release of cGAMP within APCs [327]. This targeted delivery promotes the production of IFNs and inflammatory factors, effectively bridging localized RT with robust systemic adaptive immunity.

The immunosuppressive TME remains a critical bottleneck limiting the efficacy of conventional RT and ICD. Nanomaterials offer innovative strategies to reshape this hostile milieu and synergistically amplify ICD-induced immune responses. Specifically, nanoradiosensitizers directly modulate the functional state of tumor-associated immune cells. For instance, THN combined with RT elicits ICD to activate CD⁸⁺ T cells while polarizing M2 macrophages into the M1 phenotype [187]. This transition elevates pro-inflammatory cytokines, such as IL-1β and TNF-α, and downregulates immunosuppressive factors including interleukin 10 (IL-10) and transforming growth factor beta (TGF-β). Similarly, bismuth based MOFs remodel M2 macrophages and decrease TGF-β levels, reversing TME mediated immunosuppression [95].

Additionally, certain nanoplatforms facilitate in situ immune checkpoint blockade. The ZnO–Au@mSiO₂ system releases Zn²⁺ to activate DNAzyme, which specifically degrades PD-L1 mRNA to disrupt the PD-1/PD-L1 axis [198]. Nanomaterials also enhance antigen presentation by modulating innate immune signals. HfMOF-PEG-FA upregulates toll-like receptor 7 and delivers imiquimod to activate the antigen presentation machinery, thereby potentiating antitumor immunity and RT efficacy [328].

A major limitation of ICD-based radiosensitization is the difficulty in distinguishing bona fide immunogenic cell death from mere necrotic cell death, as both release DAMPs but only ICD triggers a productive adaptive immune response. Furthermore, most studies use syngeneic mouse models that do not fully recapitulate the heterogeneity of human tumor immune microenvironments, and successful induction of ICD in animal studies has often failed to translate into clinical abscopal effects.

Translational challenges and clinical prospects of emerging pathways

Despite the promising radiosensitization potential of reprogramming tumor cell death, several critical translational barriers remain. First, the systemic toxicity of nanomedicines designed to trigger these pathways is a major concern, as the off-target induction of ferroptosis or cuproptosis in healthy organs could lead to severe side effects. Second, achieving high biological specificity is difficult due to the metabolic heterogeneity of human tumors, where the required substrates for emerging mechanisms may not be uniformly expressed. Furthermore, most findings are based on simplified animal models, which fail to replicate the complex vascular and immune environment of human malignancies. Addressing these in vivo validation challenges requires long-term biosafety assessments to ensure these innovative strategies can successfully transition from bench to bedside.

Clinical trial

Clinical evaluation of leading nano‑radiosensitizers

Promising preclinical outcomes have propelled nano-radiosensitizers into clinical pipelines, with NBTXR3 and AGuIX emerging as the most advanced candidates. NBTXR3, composed of a crystalline hafnium oxide (HfO₂) core, became the first-in-class radio-enhancer to receive European CE certification in 2019, following its success in treating locally advanced soft tissue sarcoma (STS) [110, 329]. Similarly, AGuIX, a sub-5 nm nanoparticle comprising a polysiloxane matrix with covalently grafted gadolinium chelates, offers a dual-functional profile: potential radiosensitization and superior MRI T1-contrast enhancement [330].

Clinical evaluation of NBTXR3 has demonstrated a generally acceptable safety profile and preliminary efficacy in select malignancies. A Phase I dose-escalation study (NCT01946867) in elderly patients with head and neck squamous cell carcinoma confirmed its safety, as the maximum tolerated dose was not reached and no dose-limiting toxicities were observed [331]. When combined with anti-PD-1 immunotherapy (NCT03589339), NBTXR3 yielded a 40% objective response rate in advanced solid tumors, suggesting that RT-induced ICD may help remodel the immunosuppressive TME [332]. Although trials in rectal cancer (NCT02465593) reported a 100% disease control rate and high R0 resection rates, improvements in the pathological complete response were relatively modest [333]. In a pivotal randomized Phase II/III trial (NCT02379845) for STS, NBTXR3 was associated with a higher pCR rate (16%) compared to RT alone (8%, p = 0.044) [110]. Long-term follow-up further indicated that NBTXR3 maintains a favorable benefit-risk ratio without compromising health-related quality of life [329].

As a theranostic gadolinium-based nanoparticle, AGuIX has also shown significant clinical potential. A Phase Ib trial (NCT04881032) in glioblastoma established a recommended Phase II dose of 100 mg/kg, with MRI confirming selective intratumoral accumulation (mean concentration: 84.9 µM) [334]. In the context of brain metastases, Phase I and II trials (NCT02820454, NCT04899908) demonstrated safe, selective uptake, providing imaging-based evidence for personalized dosing [335, 336]. Furthermore, in cervical cancer (NCT03308604), AGuIX provided proof-of-concept for “dose modulation” strategies, reporting a 15% sensitization effect in both external beam radiotherapy and image-guided brachytherapy, which was calculated to increase the local control probability by 3.7% [337].

In summary, while NBTXR3 and AGuIX have demonstrated acceptable safety and encouraging efficacy in early-phase trials, their full-scale clinical translation hinges on improved patient stratification, standardized administration protocols, and extended longitudinal evaluations to advance the commercialization of these intelligent nano-radiosensitizers (Table S1 and S2). It should be noted that most clinical data derive from small cohorts or single-arm studies, and larger randomized trials are needed to confirm the observed benefits.

Translational barriers limiting the broader clinical pipeline

The clinical experience with NBTXR3 and AGuIX suggests that design simplicity and predictable pharmacokinetics may be key drivers of successful translation. NBTXR3 appears to succeed through a purely physical mechanism and direct intratumoral injection, which can largely bypass systemic biological barriers, whereas AGuIX leverages its ultrasmall sub-5 nm size to facilitate efficient renal clearance and potentially reduce long-term toxicity concerns [330, 338–340]. However, the broader clinical pipeline can be hindered by interconnected obstacles, especially for more complex or systemically delivered nanostructures.

First, the potential for chronic accumulation of non-biodegradable nanoparticles in the liver and spleen remains a biosafety concern, as long-term subchronic toxicities may be insufficiently documented in preclinical studies [341]. Second, because efficient renal clearance is largely limited to nanoparticles below approximately 5.5 nm, larger platforms may experience longer retention in the reticuloendothelial system, which can complicate clinical dosimetry [342]. Third, the synthesis of multi component nanoplatforms under good manufacturing practice conditions presents practical challenges regarding batch-to-batch reproducibility and quality control [343]. Fourth, ambiguous regulatory classifications between drugs, devices, and combination products create uncertainty in approval pathways [343]. Finally, a notable hurdle exists in the quantitative dose scaling from animal models to human patients [344]. These challenges highlight the need for standardized preclinical-to-clinical translation guidelines, as well as more rigorous long-term safety data in large animal models before initiating first-in-human trials. Overcoming these multi-faceted challenges will likely require a paradigm shift from increasing structural complexity to prioritizing the biological logic of dose scaling and regulatory compatibility.

Conclusion and future perspectives

Nanomaterial-based radiosensitization strategies have evolved from a linear physical energy deposition model into a multidimensional synergistic framework encompassing physical, chemical, and biological mechanisms [345]. This review elucidates how nanoagents overcome radioresistance by inducing diverse cell death modes and remodeling the tumor immune microenvironment. However, translating these cutting-edge findings will likely require key breakthroughs in mechanistic depth, structural design, and clinical implementation.

Modern radiosensitizers act as pleiotropic regulators of biological systems rather than mere physical dose enhancers. High-Z nanomaterials not only enhance energy deposition but also trigger a series of biological cascade reactions through the SEs and ROS they generate. These reactions include inhibition of DNA damage repair, interference with cell cycle checkpoints, disruption of metabolic networks, and remodeling of the immune microenvironment [346]. This shift towards systemic biological regulation represents a fundamental paradigm change in radio-nanomedicine.

Importantly, nano agents-mediated tumor cell death reprogramming, including, but not limited to ferroptosis, cuproptosis, pyroptosis, and disulfidptosis, likely constitutes an interconnected network rather than isolated events [347, 348]. For example, ferroptotic iron overload promotes mitochondrial PT pore opening to trigger necroptosis, while mitochondrial ROS can redirect Gasdermin-mediated pyroptosis toward necroptotic signaling [347]. The convergence of DAMPs from these pathways provides a theoretical foundation for multifunctional nanoplatforms to overcome apoptotic escape [348]. Nevertheless, direct in vivo evidence demonstrating the functional significance of such cross-talk among multiple death pathways during radiosensitization remains limited.

Furthermore, nanomedicine has transformed RT from localized control into a potential engine for systemic antitumor immunity. By enhancing ICD, activating cGAS-STING, and promoting M1-type macrophage polarization, nanosensitizers convert RT into an “in situ vaccine” to potentially elicit the abscopal effect. However, immune activation remains a “double-edged sword” where excessive responses may trigger autoimmune toxicity or cytokine storms [88]. Crucially, the baseline immune milieu varies by tumor type; “hot tumors” may require direct activation, whereas “cold tumors” may necessitate microenvironmental remodeling before an immune response can be elicited [349, 350]. Looking forward, the rational combination of nano-radiosensitizers with immunotherapy represents a high-priority research direction. Rather than empirical co-administration, future studies should explore mechanistically optimized sequencing and dosing schedules: for example, determining whether STING-agonizing nanoplatforms should precede or follow immune checkpoint blockade. The development of single-particle systems that integrate radiosensitization with immune-modulatory cargos, such as anti-PD-L1 antibodies or TGF-β inhibitors, may be essential to maximize therapeutic synergy. Parallel with these immunological considerations, intratumoral spatiotemporal heterogeneity remains a major obstacle limiting RT efficacy [351]. Gradients in oxygenation, metabolism, and immune distribution within a single tumor necessitate a multidimensional approach. Although current research has developed smart-release nanomaterials responsive to TME features (such as acidic pH [177], high GSH [149], and hypoxia [352]), they primarily target differences between tumor and normal tissues rather than internal heterogeneity. Future designs should prioritize “spatially programmed” precision interventions. By integrating multimodal imaging, real-time monitoring of nanoparticle distribution could match the RT dose field, enabling dose modulation based on concentration. The quantification of intratumoral gadolinium via MRI T1 mapping in AGuIX trials exemplifies a preliminary step toward this goal [334].

Building on the need to address heterogeneity, the era of personalized nanomedicine is beginning to reshape radiosensitizer design. Current evidence indicates that nanoparticle accumulation and cell death susceptibility vary substantially across patients. Future research should aim to develop biomarker-guided patient stratification strategies, using circulating tumor DNA or imaging-based assessments of EPR capability, to identify individuals most likely to benefit from specific approaches. This precision framework could guide the selection of nanoplatforms matched to the tumor’s predominant resistance mechanism, whether it be hypoxia, apoptotic blockade, or immune exclusion.

In parallel, artificial intelligence (AI) and machine learning are poised to accelerate development at multiple levels. AI-assisted approaches could enable in silico prediction of nanoparticle biodistribution and clearance kinetics based on physicochemical descriptors, thereby reducing the experimental burden of iterative optimization. Furthermore, integrating multi-omics data with high-throughput virtual screening may identify optimal nanomaterial compositions capable of engaging multiple death modalities while minimizing off-target effects. Establishing publicly accessible databases that curate nanomaterial properties alongside in vivo efficacy and toxicity data would be vital for developing robust AI models for clinical translation.

Despite the clinical validation of NBTXR3 and AGuIX, bridging the gap between the laboratory and the clinic faces severe challenges. Precise control of physicochemical properties is essential, yet achieving batch-to-batch reproducibility remains difficult [353]. While microfluidic technology enhances uniformity, high-throughput screening methods for property optimization are still lacking [354]. Furthermore, intratumoral injection is limited by tumor accessibility, whereas intravenous administration faces rapid clearance by the reticuloendothelial system and potential systemic toxicity [353]. Current toxicological evaluations must also evolve beyond acute studies to systematically investigate long-term metabolism and potential genotoxicity [355, 356]. Overcoming these bottlenecks requires a multidisciplinary ecosystem integrating materials science, radiobiology, and immunology to truly unlock the clinical potential of nano-radiosensitization technology.

In summary, the field of nano-radiosensitization stands at a pivotal juncture. The convergence of programmable cell death modulation, patient-specific biomarker stratification, and AI-accelerated design holds considerable potential to move beyond the one-size-fits-all paradigm. Realizing this vision, where the deep integration of nanobiotechnology with radiotherapy propels oncology into a new era of precision, would require sustained interdisciplinary collaboration and a deliberate shift toward clinically translatable, scalable, and personalized nanomedicine strategies.

Author contributions

X.Y., Y.T., and Y.M. contributed equally to this work. X.P., Z.W., and Z.L. conceptualized and supervised the study. X.Y., Y.T., and Y.M. performed the literature search, analyzed the data, and wrote the original manuscript draft. L.Y. and W.W. contributed to the literature search, data curation, and manuscript revision. Z.L., Z.W., and X.P. provided critical revisions to the manuscript and acquired funding. All authors reviewed and approved the final manuscript.

Funding

The work was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (2023ZD0503000, 2023ZD0503004), the Regional Innovation and Development Joint Fund Key Project of the National Natural Science Foundation of China (U24A20735), the National Natural Sciences Foundation of China (82473434), the Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China, the Institutional Joint Innovation Fund from Sichuan University and Nuclear Power Institute of China (1-kJ-FWHT-WU-20240150), Sichuan Science and Technology Program (2025YFHZ0087), Science and Technology Project of Sichuan Provincial Health Commission (Clinical Research Special Project JH2023082), the International Science and Technology Cooperation Program of Chengdu Science and Technology Bureau (2024-YF06-00011-Hz and 2022-GH03-00004-Hz), the Health Research Project of Chengdu Eastern New Area Management Committee (202304), and 1.3.5 project for disciplines of excellence from West China Hospital of Sichuan University (ZYYC23006). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.