Radiotherapy traces its origins back to the late 19th century with the discovery of X-rays (Pfeiffer et al., 2020). Early two-dimensional (2D) planning progressed to three-dimensional (3D) conformal techniques that better protected healthy tissues. Today, IMRT is the standard for EBRT, using adjustable beam intensities to precisely target tumors. Other advances like VMAT, Image-Guided Radiation Therapy (IGRT), and Stereotactic Body Radiation Therapy (SBRT) further improve the precision of photon beam therapy (Mancuso et al., 2012; Franzone et al., 2016; Teoh et al., 2011). Proton therapy adds another dimension through the Bragg peak effect, which concentrates radiation at tumor sites while sparing deeper healthy tissues (Yan et al., 2023; Huff, 2007). While IMRT and proton therapy have improved dose conformity, their limitations are multifaceted. Photon therapy with low-energy beams (6–20 MeV) risks damaging tissues beyond the tumor due to exit doses (Galloway et al., 2012). High-energy photons (>20 MeV) reduce skin exposure but introduce neutron contamination (0.1–0.5 Sv/Gy for 18 MeV linear accelerators), complicating long-term risk assessments (Paganetti et al., 2021). Proton therapy minimize exit doses but faces variable biological effectiveness (RBE range, 1.05–1.7), leading to unpredictable tumor coverage (Traneus and Oden, 2019). Moreover, proton facilities cost over $100 million, making them accessible to <1% of patients globally (Yan et al., 2023).

The efficacy of radiotherapy is often limited by the phenomenon of radioresistance, which refers to a tumor’s ability to withstand radiation exposure. Extensive research has been conducted to uncover the mechanisms underlying cancer radioresistance. Generally, tumor radioresistance stems from hypoxia, cancer stem cells (CSCs), the tumor microenvironment, enhanced DNA repair, and various signaling pathways (Busato et al., 2022; Deng et al., 2023). Hypoxic regions (pO₂ < 10 mmHg) exhibit 3-fold higher radioresistance, prompting the use of catalytic NPs like MnO₂ to convert tumor H₂O₂into O₂, boosting oxygenation significantly in preclinical models (Pi et al., 2023). However, current NP designs predominantly target bulk tumor cells, neglecting CSC niches that cause recurrence. For example, glioblastoma CSCs (CD44⁺/CD133⁺) show 2-fold higher post-RT survival (Tsai et al., 2021), but AuNPs conjugated with DNA fragments reduce their survival significantly, highlighting the potential of CSCs-targeting strategies (Kunoh et al., 2019). Future NPs should incorporate hypoxia-responsive drug release and CSC-specific ligands to address these layered resistance mechanisms. Additionally, precision medicine based on the individual genetic and molecular characteristics is being explored as a potential approach to combat radioresistance.

Theoretically, tumors and normal tissues exhibit differential sensitivity to radiation. The radiosensitivity of highly metabolic or functionally active tumor cells usually surpasses that of adjacent normal tissues (Liu et al., 2025; Alkotub et al., 2025). The objective of radiotherapy is to maximize the irradiation dose to the tumor while minimizing exposure to surrounding healthy tissues. Therefore, balancing efficacy and safety remains critical, as healthy tissue toxicity constrains dose escalation. Experiments demonstrate that tumor-specific accumulation of NPs can achieve favorable tumor-to-normal tissue concentration ratios to minimize collateral damage (Reda et al., 2020). The size, shape, and surface functionalization of NPs critically influence biodistribution and clearance. As reported, smaller AuNPs with diameter of 1.9 nm exhibit efficient renal clearance and low systemic toxicity in murine models (Reda et al., 2020). Additionally, localized intratumoral injection of NPs further reduces systemic exposure and associated normal tissue risks (Bai et al., 2020). While high-Z NPs enhance radiation-induced DNA damage and ROS production in tumors, their catalytic activity and prolonged retention in healthy tissues could exacerbate normal cell toxicity (Gerken et al., 2023). Promising strategies include combining NPs with tumor-specific drugs to suppress DNA repair pathways, or designing biodegradable NPs to boost radiation effects while ensuring rapid metabolic clearance. Future research should prioritize NP design innovations, such as size-tunable architectures, tumor microenvironment-responsive coatings, and hybrid systems, which could maximize tumor-selective radiosensitization while mitigating collateral damage to normal tissues. For instance, hyaluronidase/ROS cascade-responsive systems demonstrate size/charge switching from ∼150 nm to sub-50 nm particles upon entering acidic tumor regions, enabling deep tissue penetration while maintaining circulatory stability (Shi et al., 2024b), while gelatin-based platforms achieve tumor-selective size transitions through pH/enzyme dual-responsive mechanisms (Khandal et al., 2025). Hybrid designs integrating metal cores with stimuli-responsive polymer coatings liked HA-modified systems may optimize both radiation dose enhancement and biological targeting precision, as demonstrated in tumor-penetrating nanocomplexes that leverage size modulation to overcome stromal barriers (Khandal et al., 2025; Shi et al., 2024b). These innovations align with emerging strategies that couple physical radiosensitization with microenvironmental adaptation (Mansouri et al., 2023; Fu et al., 2025).

Emerging technologies like FLASH-RT and MRI-guided RT offer synergistic potential. FLASH-RT’s ultra-high dose rates (40 Gy/s) protect normal tissues but require NPs stable under extreme conditions (Shen et al., 2024). Preliminary data show TaNPs enhance FLASH efficacy without compromising tissue protection (Meng et al., 2023). MRI-guided RT with GdNPs for real-time tumor tracking demonstrates superior antitumor performance without systemic toxicity or long-term side effects (Sun et al., 2020). However, progress is slowed by insufficient collaboration between radiation oncologists, nanotechnologists, and biologists. Joint efforts should focus on developing NP-RT platforms validated in both preclinical and clinical settings. Future NPs need smart features like pH/ROS-activated drug release and designs meeting safety regulations for immune impacts and long-term toxicity. For instance, pH-responsive hybrid micelles combining inorganic NPs with pH-sensitive amphiphiles enable tumor-acidic-triggered drug release, minimizing off-target toxicity while leveraging high-Z elements (Au, Bi) for localized radiation dose amplification (Moloudi et al., 2023; Gimeno-Ferrero et al., 2024). Recent advancements in pH-stabilized Prussian blue-based nanocomposites further illustrate how acid-triggered structural transformations synchronize drug release with radiosensitization, addressing dynamic TME challenges (Shi et al., 2024a). Additionally, lipid NPs functionalized with pH-responsive bicontinuous cubic phases achieve tumor-selective drug delivery, exemplifying translational designs that align with safety regulations for immune compatibility (Rajesh et al., 2022). These innovations highlight the dual utility of pH-activated mechanisms in optimizing high-Z nanodrugs for precision radiotherapy.

High-Z NPs enhance radiotherapy through physical, chemical, and biological mechanisms, each contributing uniquely to dose enhancement (Figure 1). Under X-ray irradiation, high-Z NPs can trigger a series of physical process, including the Photoelectric effect, Auger effect, and Compton scattering (Chen et al., 2019). These interactions lead to the generation of secondary electrons, X-rays, and fluorescent light, increasing energy deposition in tumors. Both simulations and experimental studies confirm that the dose enhanced factor (DEF) depends on radiation energy. The strongest effects occur at kilovoltage (KeV) X-ray energies (50–300 KeV), where the photoelectric effect dominates (Schuemann et al., 2016; Her et al., 2017). For example, 15 nm AuNPs combined with 100 KeV X-rays achieved DEF values exceeding 2.0 in breast cancer cells, whereas megavoltage (MeV) beams yielded significantly lower DEF (1.2–1.5) due to reduced photoelectric interactions (Tudda et al., 2022). Additionally, the size and concentration of AuNPs further modulate DEF, with smaller particles and intratumoral accumulation showing superior sensitization in preclinical models (Zhang et al., 2012). However, physical effects alone cannot fully explain in vivo therapeutic gains, requiring integration with chemical and biological mechanisms.

High-Z NPs enhance radiation by catalytically increasing ROS generation. High-Z NPs amplify water radiolysis to produce hydroxyl radicals (·OH), superoxide anions (O₂ ⁻), and hydrogen peroxide (H₂O₂) (Gerken et al., 2023). AuNPs, for example, produce low but biologically significant ROS levels detected by FLIM-ROX, a sensitive imaging method capable of tracking ROS in living systems (Balke et al., 2018). This ROS amplification disrupts redox homeostasis, exacerbating DNA damage, mitochondrial dysfunction, and apoptosis, thereby sensitizing cancer cells to radiation (Yadav et al., 2024; Yang et al., 2025). However, ROS quantification remains challenging due to methodological limitations. Common probes like H₂DCF-DA suffer from artifacts such as auto-oxidation, photo-bleaching, and poor specificity, leading to overestimated ROS levels (Balke et al., 2018). Electron spin resonance (ESR) detects radicals specifically but lacks sensitivity in biological systems and requires complex sample preparation, limiting its applicability in dynamic cellular environments (Yang et al., 2025). Advanced tools like multiphoton FLIM-ROX enable high-resolution, real-time ROS mapping with minimal interference (Balke et al., 2018). However, standardized protocols for ROS quantification are still needed, with emerging techniques like surface-enhanced Raman spectroscopy (SERS) improving accuracy and reproducibility (Chen et al., 2025). SERS offers distinct advantages for ROS quantification, including ultrahigh sensitivity for detecting transient ROS at low concentrations in real time (Yang et al., 2025), and the ability to achieve spatially resolved monitoring of ROS dynamics within subcellular compartments through localized plasmonic enhancement (Ding et al., 2025b; Chen et al., 2022). Unlike conventional methods, SERS minimizes interference from complex biological matrices by leveraging molecular fingerprint specificity and enzyme-mimicking signal amplification strategies (Wu et al., 2024), while recent advances in substrate engineering have significantly improved reproducibility for quantitative analysis (Kao et al., 2025; Cai et al., 2023). These features position SERS as a transformative tool for standardizing ROS measurements in radiobiological studies.

The biological mechanisms underpinning high-Z NPs-mediated radiosensitization extend far beyond physical and chemical effects. Emerging evidence highlights their ability to disrupt DNA repair pathways, modulate cell cycle progression, induce bystander effects, and reprogram the tumor immune microenvironment (Yuan et al., 2024; Das, 2025; Xu et al., 2024). For example, platinum-based NPs (PtNPs) have been shown to inhibit BRCA1-mediated DNA repair, worsening radiation-induced DNA damage (Hullo et al., 2021). Gadolinium-carbon dots (Gd@Cdots) trap cells in the radiation-sensitive G2/M phase, amplifying chromosomal fragmentation and mitotic catastrophe (Lee et al., 2021). Intriguingly, AuNPs can trigger bystander effects through the release of mitochondrial ROS and pro-apoptotic factors, sensitizing neighboring cells without direct NP uptake (Choi et al., 2018). Emerging evidence also implicates significance of immune modulation. Iridium (Ir)-based nanoplatforms polarize tumor-associated macrophages toward the pro-inflammatory M1 phenotype and promote dendritic cell maturation, fostering systemic antitumor immunity alongside localized radiosensitization (Zou et al., 2023). Moreover, combinatorial strategies utilizing high-Z NPs with immunogenic cell death (ICD) inducers enhance antigen presentation and T-cell infiltration, overcoming radioresistance in immunologically “cold” tumors (Zhen et al., 2023). While biological sensitization offers a multidimensional approach to radiotherapy enhancement, its clinical translation requires addressing NP heterogeneity in cellular uptake and off-target immune activation.

The conventional “three-phase” model (physical-chemical-biological) fails to capture systemic effects like immune modulation or bystander effects. A “systems radiobiology” approach integrating multi-omics is needed to map NP-induced molecular networks (Karapiperis et al., 2021). Furthermore, while most research focuses on photon beams, NP interactions with protons or carbon ions remain under explored. Monte Carlo simulations suggest Au, Pt, Gd, and Fe NPs enhance proton energy deposition by 14–27% at 5–50 MeV (Martinez-Rovira and Prezado, 2015; McKinnon et al., 2016), but experimental validation is lacking. To translate these mechanisms into clinical practice, NP formulations should be co-engineered with imaging tracers for real-time monitoring of tumor distribution and dose enhancement, which is a critical step toward personalized radiotherapy.

The radiosensitization efficacy of high-Z NPs such as Au, Gd, and Bi is intrinsically linked to their physicochemical properties, including surface modifications, morphology, and size. GNPs have been extensively studied due to their tunable size (2–100 nm) and morphology-dependent optical properties, which enable precise control over surface plasmon resonance effects (Ding et al., 2025a; Kaercher and Lear, 2025). For instance, thiolated surface functionalization enhances GNPs’ biocompatibility and reduces aggregation in biological environments (Wang et al., 2025), while ATP-coated ultrasmall GNPs (2–5 nm) exhibit selective binding capabilities to biomolecular targets (Katrivas et al., 2023). The surface roughness of GNPs, influenced by aspect ratios (1:1 to 1:10), also plays a critical role in minimizing cytotoxicity when integrated into biomedical devices (Shin et al., 2023). In contrast, Gd-based nanoparticles, such as gadolinium oxide (GdO), demonstrate unique advantages due to their paramagnetic properties and high X-ray attenuation. Surface modification with bovine serum albumin (BSA) in GdO@BSA-Au hybrid NPs not only improves colloidal stability but also synergistically enhances radiation dose deposition within tumors (Nosrati et al., 2023). Spherical Gd orthoferrite NPs with uniform surface morphology further exhibit enhanced biocompatibility and cellular uptake, as confirmed by FE-SEM and HR-TEM analyses (Guganathan et al., 2022). Bismuth oxide (BiO) NPs, on the other hand, are notable for their high photoelectric absorption cross-section. Surface functionalization with β-cyclodextrin (β-CD) improves their dispersibility and enables efficient drug loading (Alex and Mathew, 2023), while BiSe nanosheets with thiolated gold nanoclusters achieve superior surface area and charge modulation for targeted applications (Li et al., 2024). Size-dependent effects are particularly pronounced in Bi NPs, where smaller particles (<30 nm) exhibit higher radiosensitization due to increased surface-to-volume ratios (Guruswamy Pandian et al., 2025). Importantly, Monte Carlo simulations reveal that secondary electron emission peaks at distinct energies for Au (30 keV), Bi (30 keV), and Gd (60 keV), highlighting material-specific radiation enhancement mechanisms (Mansouri et al., 2024). Collectively, these findings underscore the necessity of tailoring surface chemistry, size, and morphology to optimize the therapeutic index of high-Z NPs, with GNPs excelling in tunable surface engineering, Gd NPs in multimodal imaging compatibility, and Bi NPs in high-Z-driven dose enhancement (Stergioula et al., 2023). Future research should focus on standardizing synthesis protocols to reconcile disparities in reported dose enhancement factors across studies.

Survival Fraction (SF) is a fundamental radiobiological parameter quantifying the proportion of cells retaining clonogenic potential after irradiation. Derived from clonogenic assays (Subiel et al., 2016), SF reflects the cumulative effects of DNA damage and repair mechanisms, serving as a critical endpoint to evaluate radiation efficacy and sensitizer performance (Figure 2). The linear-quadratic (LQ) model (SF = exp (-αD-βD²)) distinguishes between lethal (α) and sublethal (β) damage (Hong et al., 2024; Abdollahi et al., 2024). Higher α values indicate stronger radiosensitization by increasing irreparable damage per unit dose. For example, thulium (III) oxide NPs (Tm₂O₃) combined with carboplatin reduced SF in metastatic cutaneous squamous cell carcinoma models compared to radiation alone (Perry et al., 2020). However, recent studies highlight emerging insights into the biocompatibility and toxicological profiles of TmO NPs. For instance, TmO NPs designed with varying thulium compositions demonstrated enhanced X-ray absorption and ROS generation capabilities, suggesting oxidative stress as a potential mechanism of toxicity (Zhu et al., 2025b). Additionally, critical gaps persist in understanding organ-specific accumulation and chronic exposure effects of TmO NPs, necessitating systematic in vivo toxicokinetic studies to bridge current knowledge limitations. Similarly, gold nanowires also suppressed SF in breast cancer models more effectively than nanospheres, likely due to higher oxidative stress and α elevation than spherical counterparts (Bai et al., 2020). Notably, TiO₂ nanotubes were shown to increase the α value while simultaneously decreasing the β value (Mirjolet et al., 2013). These findings suggest a higher α/β ratio in tumor cells treated with nanoradiosensitizers, indicating increased tumor sensitivity to ionizing radiation. Basically, most studies use a 2 Gy dose to evaluate the effectiveness of NPs in vitro, as it corresponds to the standard dose per fraction in conventional radiotherapy (Subiel et al., 2016). However, some studies calculate the DEF based on survival levels using acute doses of 3 Gy (SF₃) (Coulter et al., 2012; Taggart et al., 2014), 4 Gy (SF₄) (Taggart et al., 2014; Maggiorella et al., 2012), or 8 Gy (SF₈) (Taggart et al., 2014; Maggiorella et al., 2012). Table 1 listed nanoparticle studies using survival fraction to calculate dose enhancement effect (Table 1). While SF remains a gold standard, nanoparticle off-target effects may complicate clonogenic assay results. Innovations in 3D tumor models and real-time SF monitoring could refine predictive power, bridging in vitro findings to clinical translation.

The mean inactivation dose (MID), calculated as the area under the SF curve, reflects the average dose required to inactivate a cell population (Figure 2A). The concept of MID was firstly introduced to assess survival curves of mammalian cells. Recent advances in high-Z NP-mediated radiosensitization used MID to quantify dose enhancing effects. For example, studies using iron oxide NPs (IONs) under proton irradiation demonstrated enhanced localized energy deposition via Coulomb nanoradiator (CNR) effects, which amplified secondary electron emission and ROS generation by 1.2 to 2.5 fold (Jeon et al., 2016). Diverging from conventional metrics, MID offers a holistic view of cellular response dynamics, particularly valuable for high-Z NPs where heterogeneous energy deposition complicates traditional models. Table 2 listed NP studies using MID to calculate dose enhancement effect (Table 2).

The dose enhancement factor (DEF), defined as the ratio of radiation dose in the presence of NPs to that without NPs (Figure 2B), primarily quantifies the physical dose amplification from the high photoelectric and Auger electron yields of high-Z materials (Chow and Jubran, 2023). In contrast, sensitizer enhancement ratio (SER), calculated as the ratio of radiation doses required to achieve the same biological effect with and without NPs (Figure 2C), incorporates both physical dose enhancement and biological interactions (Guerra et al., 2022). Studies on high-Z NPs have demonstrated significant DEF and SER values under varying irradiation conditions. For example, AuNPs exhibited a DEF of 5.7–8.1 in clinical megavoltage (MeV) beams at depths up to 30 cm (Gerken et al., 2024). Similarly, magnetic FeO@AuNPs achieved a DEF of 22.17% in cytoplasm under a magnetic field, demonstrating the role of NP targeting in enhancing local dose deposition (Mesbahi et al., 2022). However, SER values are more context-dependent. For glioblastoma cells, AuNPs achieved SER values of 1.5–1.8, while iron oxide NPs (IONPs) showed lower SER values (1.09–1.32), emphasizing the importance of NP composition and cell type (Guerra et al., 2022). Notably, coating layers and aggregation states of NPs also influence DEF and SER, as thicker coatings may attenuate secondary electron emission, while optimized surface functionalization can improve tumor retention and radiation interaction (Mansouri et al., 2023).

A critical perspective emerges from the interplay between DEF and SER. While DEF often dominates in kilovoltage (KeV) X-rays due to strong photoelectric effects, but SER gains importance in MeV beams through biological mechanisms like ROS amplification (Gerken et al., 2024). For example, Hf-based NPs achieved a SER of 1.55 at 30 cm depth under MeV beams, suggesting that biological sensitization may outweigh physical dose enhancement in deep-seated tumors (Gerken et al., 2024). However, challenges persist in translating these metrics to clinical practice. Variability in NP distribution within tumors, inconsistent DEF-SER correlations across studies, and the lack of standardized protocols for measuring these parameters hinder robust comparisons. A summary of DEF and SER used in NP studies can be found in Table 3 (Table 3).

ROS generation and DNA damage are crucial to NP-mediated radiosensitization. However, their quantification faces technical challenges. Fluorescent probes like H₂DCF-DA are widely used for ROS detection due to their accessibility and compatibility with live-cell imaging. However, limitations such as auto-oxidation, photobleaching, and interference from intracellular thiols or metal ions often lead to false positives or underestimation (Stergioula et al., 2023). For example, studies involving AuNPs found differences in ROS quantification when comparing H₂DCF-DA with ESR, which has lower sensitivity, requires higher sample-volumes, and cannot resolve spatial-temporal dynamics in biological systems (Stergioula et al., 2023). Similarly, DNA damage quantification relies heavily on γ-H₂AX foci imaging, a marker for DSBs (Figures 3A,B). This method is semi-quantitative and accessible, but cannot distinguish between direct radiation-induced damage and NP-specific chemical interactions, potentially overestimating therapeutic efficacy (Bemidinezhad et al., 2024).

Recent studies show that high-Z NPs enhance both ROS and DNA damage (Figures 3C,D). For example, Au@AgBiS core-shell NPs increased ROS generation by 2.3 times compared to radiation alone, as validated by fluorescent probes and ESR (Xiao et al., 2023). Hafnium-based NPs like NBTXR₃ increased DSBs by 40% in melanoma models, quantified by γ-H₂AX foci and comet assays, and correlated with better tumor control in preclinical trials (Zheng and Sanche, 2023). However, inconsistencies arise when linking in vitro quantification to clinical outcomes. For example, superparamagnetic iron oxide NPs (SPIONs) showed persistent γ-H₂AX foci in melanoma cells, indicating unrepairable DNA damage, but their clinical translation is limited due to unresolved toxicity and off-target effects (Bemidinezhad et al., 2024).

In vivo evaluation of high-Z NP radiosensitization uses various metrics to measure survival benefits and tumor response (Table 4). Median survival time (MST) is a key endpoint, indicating survival benefits and often correlated with tumor control and treatment durability. For example, glioblastoma-bearing mice treated with PEGylated-gold nanoparticles had an MST of 28 days, compared to 14 days in controls, showing the radiosensitizing potential of targeted NPs (Figure 4A) (Yang et al., 2022). mAuNPs alone showed no improvement in survival of B16-F10 cell-bearing mice (16 days, similar to PBS controls), combined therapy with carbon ion irradiation extended survival to 42 days (Figure 4B). In HER3-expressing tumor models, Z-ABD-Z-mcDM1 conjugates extended MST from 68 days for monotherapy to 90 days by enhancing radiation-triggered drug release (Zhang et al., 2024). Similarly, Lu-FAP-2287, a lutetium-based radiopharmaceutical, combined with anti-PD-1 immunotherapy, suppressed tumor growth and extended survival in fibrosarcoma models (Chen et al., 2024). Tumor volume tripling time (TVTT), another metric, reflects regrowth kinetics and provides additional insights. Studies on pancreatic cancer revealed median volume doubling times (VDT) of 40 days, with growth rates inversely correlated to tumor size (Figure 4C) (Hussain et al., 2023). Tumor growth inhibition (TGI), which measures the reduction in tumor volume compared to untreated controls, includes both cytostatic and cytotoxic effects. For example, Gd2O3@BSA-Au NPs achieved 50% TGI in 4T1 tumor-bearing mice, (Figure 4D) (Kim et al., 2024). Notably, NPs like CMP (cationic polymer-coated Au NPs) combined with irreversible electroporation (IRE) achieved over 80% TGI and extended survival in mice by enhancing immunogenic cell death (Atkinson et al., 2025). These metrics connect preclinical findings to clinical relevance, aiding dose optimization and mechanistic validation.

Although MST and TGI are prevalent in current studies, integrating tumor volume kinetics and temporal heterogeneity could enhance outcome predictions. Faster-growing tumors might need higher NP concentrations or fractionated radiation to boost TGI (Feucht et al., 2024). The correlation between NP retention time and MST highlights the importance of pharmacokinetic optimization. However, challenges like standardizing growth rate measurements and dealing with interspecies variability in NP biodistribution remain (Xu et al., 2022). Future research could investigate multimodal imaging with dynamic TGI modeling for personalized NP-enhanced radiotherapy. In summary, in vivo quantitative metrics such as MST, TVTT, and TGI are crucial for understanding the therapeutic potential of high-Z NPs. Combining these metrics with mechanistic studies and advanced imaging will speed up the development of precision radiosensitization strategies.