Functional inorganic nanoparticles in cancer: Biomarker detection, imaging, and therapy

Cancer represents a significant global public health concern. Based on GLOBOCAN 2020 data, there were ∼19.3 × 10⁶ new cancer cases and nearly 10 × 10⁶ cancer-related deaths worldwide. The projected global cancer burden is expected to reach 28.4 × 10⁶ cases by 2040.¹ With the development of medical technology and the emergence of various treatment methods and drugs, the overall cancer mortality rate has decreased by ∼32% in the past 30 years. However, the incidence of cancer cases and cancer-related deaths is still increasing rapidly worldwide.² Cancer-related deaths primarily result from the metastatic and invasive abilities of cancer cells, accounting for ∼90% of cancer-related fatalities.³ Thus, it is crucial to explore more effective early diagnosis methods and develop new and efficient treatment techniques to improve the sensitivity and specificity of cancer detection and treatment. However, existing treatments have their limitations and make it difficult to achieve the desired treatment for all types of cancer.⁴ Drug resistance and side effects are also important challenges that need to be addressed.⁵ In addition, there are still many deficiencies in the existing cancer detection technology, such as the low detection rate of early cancers and the problems of false positives and false negatives. The sensitivity and specificity of the assay are low. Some advanced technologies are costly and difficult to popularize. In recent years, nanotechnology has emerged as a promising approach for addressing these challenges and has garnered significant attention and research in various fields.^6–10 

Nanomaterials exhibit unique surface effects, small size-dependent properties, quantum size effects, and macroscopic quantum tunneling effects.^11,12 These characteristics are conducive to the construction of a diagnosis and treatment platform based on cancer characteristics.^13,14 At present, there are many nanomaterials used in diagnosis and treatment systems, mainly comprising organic nanoparticles (NPs) and inorganic nanoparticles. Organic nanoparticles (excluding carbon-based nanoparticles) encompass nanomaterials and nanostructures comprised of organic substances, such as liposomes, polymers, micelles, and dendrimers. Inorganic nanomaterials consist of metals or metal oxides, including pure precious metal nanoparticles such as gold or silver, metallic oxides such as titanium dioxide and zinc oxide, as well as semiconductors such as silicon and various ceramics.^15,16 Organic nanomaterials have good biocompatibility and stability.¹⁷ Compared with organic nanomaterials, inorganic nanoparticles possess optical, thermal, electrical, magnetic, and mechanical properties, catalytic property, a large surface area, and easy surface modification, which make them highly promising for biomedical engineering applications.¹⁸ Using the properties of inorganic nanoparticles, stable, efficient, and safe carriers can be constructed to load anti-cancer drugs or nucleic acids for cancer therapy.^19,20 The diverse structures and properties of inorganic nanoparticles enable the simultaneous delivery of drugs with different mechanisms of action, leading to synergistic effects that hold great significance for cancer treatment.²¹ For example, Li et al. constructed pH-responsive lipid-coated CaP nanoparticles (LCP NPs) and used them to co-load Cu²⁺ and disulfiram (DSF). Following intravenous injection, the nanoparticles preferentially accumulate in the cancer due to their longer blood half-life. Subsequently, in the acidic tumor microenvironment (TME), the nanoparticles degrade, releasing Cu²⁺ and DSF. This process generates a cytotoxic metabolite, the diethyldithiocarbamate–copper complex (CuET), which is effective in cancer therapy. In addition, the active metabolite CuET can also effectively induce immunogenic cell death (ICD) in cancer cells, thereby regulating the immunosuppressive TME and enhancing the systemic immune response triggered by immune checkpoint blocking (ICB) therapy.²² In addition, nanomaterials can also be used to construct high-accuracy probes for cancer diagnosis. Nanomaterials play a sensitive, accurate, rapid, and straightforward role in cancer detection and diagnosis, enabling early detection, reducing misdiagnosis rates, and alleviating patient suffering.^23,24 Inorganic nanoparticles, often containing metal elements or doped metal ions, exhibit unique properties such as x-ray absorption, magnetism, optics, and acoustics, which make them suitable for various biomedical imaging techniques such as computed tomography (CT) imaging, magnetic resonance imaging (MRI), nuclear medicine imaging, fluorescence imaging (FLI), photoacoustic imaging (PAI), and combined imaging. Importantly, when combined with radiotherapy, chemotherapy, gene therapy, immunotherapy, photothermal therapy (PTT), photodynamic therapy (PDT), and sonodynamic therapy (SDT), inorganic nanoparticles offer unique advantages in cancer treatment.²⁵ 

Unlike small molecule materials that are usually rapidly metabolized, inorganic nanoparticles, due to their relatively larger size, can accumulate in the reticuloendothelial system for extended periods. The potential toxicity of inorganic nanoparticles seriously hinders their application in clinical therapy. Moreover, the high surface energy of inorganic nanoparticles also leads to their easy agglomeration and encapsulation by proteins in body fluids, which may change their surface charges and properties.²⁶ Therefore, surface modification of inorganic nanoparticles is usually required to further optimize their functions.^27,28 Chemical modification of the surface of nanomaterials can increase their dispersion stability in water, serum, and various solvents, enhance their biocompatibility, and reduce their toxicity in biological systems.²⁹ Functional inorganic nanoparticles represent a rapidly evolving field in cancer therapy research, involving surface chemical modifications to meet the biological and clinical requirements of cancer treatment.^30,31 Functional inorganic nanoparticles not only enhance their physical properties, drug-loading efficiency, and enrichment capabilities but also fulfill the need for in vivo degradation and self-clearance (Fig. 1). For example, nanoparticles are encapsulated in shells of different materials. When exposed to certain stimuli, these particles undergo separation, expansion, dissolution, or rearrangement to change the particle size within the shell.³² Liu et al. designed a pH-stimulating-triggered drug delivery system based on hollow mesoporous silica nanoparticles (HMSNs) for cancer therapy. They introduced β-cyclodextrin (β-CD) functionalized with 3-(3, 4-dihydroxyphenyl) propionic acid onto the surface of HMSNs via a borate-catechol ester bond. Subsequently, polyethylene glycol (PEG) conjugated with adamantane was anchored to HMSNs-β-CD nanocarriers via host–guest interactions. Once the system penetrated the cancer through the enhanced permeability and retention (EPR) effect under weakly acidic conditions, the benzoic acid-imine bond between PEG and Ada was cleaved. In addition, the detachable PEG protective layer promoted cellular uptake of the HMSNs system. Subsequently, even under low pH endosome conditions, the borate–catechol ester bond linker was further hydrolyzed to deliver the drug into the cell, leading to efficient apoptosis.³³ Therefore, the development of functional inorganic nanoparticles is of great significance for cancer diagnosis and treatment. Although the application of inorganic nanoparticles in cancer therapy has been reviewed, the focus on functional inorganic nanoparticles is limited. In this Review, we describe the characteristics of several representative inorganic nanoparticles, discuss strategies for functional modification Table I, and explore the advantages of functional inorganic nanoparticles in cancer biomarker detection, imaging, radiotherapy, chemotherapy, gene therapy, immunotherapy, PTT, PDT, SDT, and combination therapy. Finally, we analyze the toxicity and potential challenges associated with functional inorganic nanoparticles. This Review provides a more comprehensive overview of the latest research progress in the application of functionally modified inorganic nanomaterials in the detection, imaging, and treatment of cancers. By integrating multiple research contents, it provides readers with a more comprehensive perspective so that readers can systematically understand the application of functionally modified inorganic nanomaterials in cancer biomedicine. In addition, this Review also summarizes the shortcomings of functionalized modified inorganic nanomaterials in cancer applications and predicts the development direction and application prospects of inorganic nanomaterials in the next few years. This Review aims to provide an overview of the application of functional inorganic nanoparticles in cancer diagnosis and treatment, opening up new possibilities in this field.

We classified inorganic nanoparticles as inorganic nonmetallic nanoparticles and inorganic metallic nanoparticles according to whether they contain metallic elements or not. Inorganic nonmetallic nanoparticles mainly refer to all kinds of nonmetallic oxides, carbon-based nanomaterials, various semiconductors such as silicon, graphene, and ceramics. Inorganic metallic nanoparticles mainly refer to inorganic nanoparticles containing metal elements, including nano-gold, nano-silver, nano-titanium, and other nano-metal materials.^16,75 The functional modification of inorganic nanoparticles can mainly include the adjustment of size, shape, and potential; surface modification of organic nanomaterials; surface modification of inorganic nanomaterials; and surface modification of functionalized components (fluorophores, radionuclides, and targeting ligands). Modifying the size, charge, or potential of inorganic nanoparticles can increase their transport and delivery effects.^32,76 Targeted molecular modification can equip nano-carriers with “navigation systems” to achieve the precise effect.^77–79 For example, stimulus-responsive-modified inorganic nanoparticles tend to aggregate and release drugs under specific conditions, achieving anti-cancer effects while minimizing damage to surrounding tissue.^80–82 Inorganic nanoparticles modified with various cell membranes (CMs) can camouflage themselves and mimic their biological effects.^27,79,83 The imaging function of inorganic nanoparticles can be used to construct platelet camouflage nanoprobes with active targeting characteristics. This biomimetic nanoprobe demonstrates the ability to evade macrophage phagocytosis while specifically binding to CD44 receptors present on the surface of most tumor cells.⁸⁴ The surface functional modification of inorganic nanoparticles not only improves their therapeutic efficacy, stability, and biosafety but also allows for combination with hyperthermia and other molecular drug therapies to enhance the effectiveness of cancer diagnosis and treatment in multiple ways.

Inorganic nonmetallic nanoparticles are widely used due to their large specific surface area, good biocompatibility, and low cost. In biomedical applications, it is mainly used as a drug delivery carrier to achieve precise targeted, improve drug efficacy, and reduce side effects. In addition, it can also be used to construct biosensors for the detection of biomarkers to provide strong support for the early diagnosis of cancer.

The uniform pores of mesoporous silica nanoparticles (MSNs) facilitate continuous and controlled drug release.^75,85,86 By encapsulating drugs within their pores and adding a protective coating, MSNs can prevent the degradation of sensitive biomacromolecules in the bloodstream.^87,88 Silica-based agents, incorporating MSNs, are widely used due to their high absorption, good dispersion, and suitable thickening properties. For example, adding a small amount of MSNs to drugs such as ranitidine and cimetidine can alter the mobility of these drugs.⁸⁹ The stable framework structure of MSNs presents challenges in terms of degradation and elimination from the body, potentially causing organ damage with long-term accumulation. MSNs have two functional surfaces: an inner surface consisting of pore channels and an outer surface that is highly susceptible to selective functionalization.^90,91 In aqueous solutions, the structure of MSNs undergoes hydrolytic decomposition of siloxane (Si–O–Si) bonds, transforming into biocompatible and excretable orthosilicic acid [Si(OH)₄]. MSNs can be combined with organo-bridged alkoxysilanes to promote their complete degradation in body fluids [Fig. 2(a)].³⁷ Moreover, MSNs can be chemically bonded to drugs through stimulus-responsive mechanisms, allowing targeted drug delivery and controlled release in response to specific stimuli.⁹² Hydroxycamptothecin and siMCT-4 were loaded on a GSH-responsive hollow mesoporous organo-silica nanoplatform, which triggered a cascade reaction in the weakly acidic TME and high GSH levels in cancer cells. This nanoplatform can effectively eliminate the immunosuppressive TME, inhibits cancer growth, and combines chemotherapy with lactate efflux inhibition.⁹³ In addition, targeted molecules can be attached to the surface of MSNs. For example, folic acid (FA) can bind to the cell membrane of cancer cells overexpressing folate receptors, which enhances the uptake of nanoparticles by cancer cells via receptor-mediated endocytosis [Fig. 2(b)].⁹⁴ Stimulating responsive inorganic nanoparticles enables precise drug release and enhances immune system recognition of cancer locations.^95,96 MSNs fixed with molecular and supramolecular switch surfaces can take advantage of the switching ability of molecular and supramolecular switches to endow smart hybrid nanomaterials with intelligent and controllable properties in response to various external stimuli, such as pH, enzymes, light, and temperature.^97,98 Supramolecular nano-valves, combined with nanoparticles, pave an advanced path for the construction of gated materials for controlling drug release.^99,100 Supramolecular nano-valves based on the synthetic macrocyclic arenes are modified on MSNs, and supramolecular nano-flap functionalized nano-supports exhibit superior properties in response to various types of internal and external stimuli. Due to the multi-functional functionalization and dynamic host–guest properties of synthetic macrocyclic aromatic hydrocarbons, mechanized nanocarriers are endowed with ideal targeting capability, efficient operation, and precise responsiveness.¹⁰¹ Yang et al. utilize a pyridine (Py)-modified HMSN nanoparticle-based drug reservoir (HMNS-PY) to form a versatile supramolecular drug delivery platform. The HMSN-Py surface is connected by supramolecular host–guest interaction with a layer of near-infrared (NIR)-operable carboxylatopillar[5]arene (CP5)-functionalized CuS (CP5-CuS) nanoparticles between the CP5 ring and the Py stalks, and another layer of Fa-conjugated polyethylene glycol (FA-PEG) antennas enables the active targeting of cancer lesions through electrostatic interactions. CP5-CuS NPs not only act as a quadruple stimulus-responsive nanogate for controlled drug release but also realize photothermal therapy. At the same time, the anti-cancer drug doxorubicin (DOX) can be released from the HMSN-Py reservoir in the cancer microenvironment for chemotherapy, enabling multimodal synergistic therapy.¹⁰² Designing stimulation-responsive inorganic nanoparticles with multiple responses that can respond to multiple triggers enables the alteration of the cancer environment in various ways, leading to damage to the cancer and weakening of its defense mechanisms. Compared to nanoparticles responsive to a single-stimulus, multi-stimuli-responsive inorganic nanoparticles offer enhanced intelligence and efficiency in controlled drug delivery.

Graphene oxide (GO) can be functionalized using surface-linked proteins, peptides, and organic small molecules.³⁹ GO can enable the drug containing π–π structure to be loaded on GO by π–π stacking.^103,104 For example, Tu et al. demonstrated the attachment of cyclic R10 peptide to polyglycerol-covered GO and the successful loading of DOX [Fig. 2(c)]. This study confirmed that the modified GO achieved more precise drug release and exhibited a favorable anti-cancer effect.¹⁰⁵ In addition, the surface of GO contains a large number of carboxyl groups, hydroxyl groups, and other functional groups, thereby simplifying the process to obtain positively charged GO complexes with covalent or non-covalent surface modifications for gene delivery.¹⁰⁶ For example, cations such as polyethylenimine, chitosan, and 1-pyrene methylamine have been used to modify GO to construct graphene-based gene delivery vectors. To expand the functionality of GO, its surface can also be decorated with various nanoparticles such as Au, Ag, Pt, Cu, MnO₂, and Fe₃O₄. Through functional modification, GO has been widely used in catalysts, contrast agents, drug carriers, and photothermal reagents.⁴⁰ 

Black phosphorus (BP) can be degraded into non-toxic and harmless phosphate in vivo, displaying excellent biocompatibility and high safety.⁴¹ In addition, BP possesses a bandgap of up to 2.0 eV, enabling absorption across the ultraviolet (UV) and visible regions. The instability of BP at room temperature largely limits its application. BP can be improved through non-covalent or covalent functionalization, involving modifications with polymers, liposomes, small molecules, or metal nanomaterials. Wan et al. modified BP using PEG through electrostatic interactions, thus enhancing its biocompatibility and physiological stability.⁴² Zhao et al. used Nile blue to covalently modify BP. Modified BP demonstrates enhanced storage stability, outstanding photothermal cancer ablation efficacy, efficient NIR imaging, and excellent biocompatibility, thus presenting a promising future in the field of anti-cancer therapy.¹⁰⁷ Moreover, surface modification can enhance the delivery capability of BP. The combination of chitosan and BP quantum dots (QDs), along with PEG surface modification, facilitates efficient penetration of the lung mucus barrier and improved drug delivery. Hydrophilic PEG and positively charged chitosan aid the nanocarriers in quickly traversing the mucus layer and adhering to epithelial cells.¹⁰⁸ 

The unique tubular structure of carbon nanotubes (CNTs) significantly increases the surface area, making them attractive as molecular drug carriers.^44,45 To address the poor dispersion of original CNTs, functionalized CNTs can be positively charged and covalently bound with amide groups, greatly improving their dispersion for loading and transporting nucleic acids. Phosphatide-coated PEG-functionalized CNTs, NH₄⁺-functionalized dendritic CNTs, and carboxylated CNTs can effectively deliver small interfering RNA (siRNA) to cells.^46,47 Zhao et al. treated the original CNTs with nitric acid to form a carboxyl modification on the surface of CNTs. Peptide lipid (PL) and sucrose laurate (SL) were utilized to form a functionalized delivery system with excellent temperature sensitivity. Moreover, PL- and SL-modified CNTs demonstrated favorable biocompatibility. In addition, delivery systems loaded with therapeutic siRNA can synergize PTT and gene therapy to achieve a strong synergistic anti-cancer effect [Fig. 3(a)].¹⁰⁹ 

Compared with rod-shaped CNTs and lamellar graphene, carbon nanohorns (CNHs) have a unique spherical morphology. Single-walled CNHs (SWCNHs) have a large specific surface area and numerous angular voids, enabling them to adsorb a large number of molecules. Moreover, they can be used as an ideal drug carrier.⁴⁸ In addition, the small pore size of SWCNHs is suitable for the adsorption of some relatively small molecules. The production process of SWCNHs does not involve the use of metal catalysts to avoid cytotoxicity from metal impurities. SWCNHs can assemble into micrometer-scale bundles or form spherical aggregates, enhancing drug permeability and retention under passive cancer-targeting conditions. This accumulation near cancer tissues improves their anti-cancer efficiency. Studies have demonstrated the adsorption and slow release of dexamethasone and cisplatin on SWCNHs.⁴⁹ Zhong et al. used chitosan-modified oxidized SWCNHs for the loading of DOX, where the aromatic ring structure of DOX and ox-SWCNHs formed reversible T–T bonds, resulting in a drug-loading efficiency of up to 60%. The use of chitosan as a modifier can significantly improve the water solubility and biocompatibility of the whole system.⁵⁰ 

The good water dispersibility of carbon dots (CDs) greatly enhances the solubility of hydrophobic small molecule drugs and promotes their release in acidic cancer environments. The abundant functional groups on the surface of CDs enable easy surface modification.⁵¹ Hou et al. modified the surface of CDs with Asp-Ala-Thr-Gly-Pro-Ala peptides, which are sensitive to fibroblast-activating proteins. They then loaded the CDs with DOX. Their results showed that the nanoplatform reduced the fibrotic stroma, enhanced the drug delivery, elicited robust anti-cancer immunity, and suppressed primary cancers and metastases [Fig. 3(b)].⁵² Moreover, CDs contain numerous π electrons that behave similarly to the free electrons of metallic nanomaterials. They selectively accumulate in cancers through the EPR effect, primarily in the red to NIR region.

Inorganic metal nanoparticles contain metallic elements with unique x-ray absorption, optical, electrical, magnetic, thermal, and mechanical properties and are often used in various biomedical imaging applications. In addition, inorganic metal nanoparticles can also be used in PTT, PDT, and SD, so as to achieve an integrated platform for imaging and treatment.

With their large specific surface area, gold nanoparticles (AuNPs) can easily bind to diverse biological macromolecules. Surface-modified AuNPs with cationic polymers can form electrostatic or covalent bonds with anionic drugs. These AuNPs have been widely used as effective carriers of siRNA and DNA enzymes.^54,110 AuNPs can connect to other molecules by forming Au–S bonds with sulfhydryl groups or by electrostatic attraction. For example, drug or gene delivery can be achieved by linking Au–S to AuNPs. Moreover, targeting groups can be connected to AuNPs to improve targeting efficiency, and stimulus-responsive groups can be connected to AuNPs to achieve different response functions.¹¹¹ The FA receptor exhibits high expression in various cancer cells. The FA-modified curcumin (CUR)-loaded Au–polyvinyl pyrrolidone nanotubes enhance the sustained release of the anti-cancer drug curcumin. Peptides exhibit a strong affinity by binding to a specific receptor's binding domain or interfering with ligand–receptor interactions. AuNPs can be precisely targeted to the nucleus and mitochondria by modifying their surfaces with different peptides, respectively.¹¹² In addition, Au itself possesses electrical, magnetic, and optical properties, which endow AuNPs with unique photothermal properties. Consequently, the photothermal properties of AuNPs have found applications in various in vivo hyperthermia applications.⁵⁵ For example, Mao et al. successfully prepared tAuNP and mAuNP by coupling 2,5-diphenyltetrazole and methacrylic acid to the surface of AuNPs, respectively. Subsequently, they adsorbed the chemiluminescence substrate luminol onto mAuNP to obtain mAuNP/Lu that could produce endogenous chemiluminescence under the catalysis of H₂O₂. This allowed for the selective aggregation of AuNPs in the TME. In vitro experiments demonstrated that mAuNP/Lu can generate strong chemiluminescence upon H₂O₂ activation, leading to the photocross-linking of tAuNP and mAuNP/Lu to form gold nanoaggregates. These aggregates induced a strong localized surface plasmon resonance (LSPR) effect and activated PAI and PTT functions.¹¹³ 

Magnetic nanoparticles encompass various materials, including Fe, Co, Ni, Mn, and their oxides. Among them, iron oxide nanoparticles (IONPs) are widely used in biomedicine due to their high magnetization, good biocompatibility, and low toxicity.^114,115 Magnetic nanoparticles are typically surface-functionalized with hydroxyl, carboxyl, and other functional groups. This functionalization is achieved through covalent, hydrophobic, electrostatic, and chelating interactions. The modification of magnetic nanoparticles directly influences their physical and chemical properties, as well as their therapeutic and delivery efficiency as agents and carriers.¹¹⁶ The addition of silica on the surface of magnetic nanoparticles improves their stability and dispersion in solution and enhances their capacity to adsorb molecular drugs for efficient delivery.^57,58 Magnetic nanoparticles undergo Brownian or Nair relaxation in an alternating magnetic field, and these relaxation processes subsequently generate heat, called magnetothermal interactions. The heat generated by these magnetic nanoparticles can change the conformation of the heat-responsive macromolecules overlying their surface, thereby releasing the drug. Thomas et al.¹¹⁷ devised a novel material that incorporates zinc-doped iron oxide nanocrystals within a mesoporous silica framework that has been surface-modified with pseudorotaxanes. When iron trioxide particles produced magnetocaloric effects, they caused the molecules that act as valves to decompose and induce drug release in the pores.¹¹⁷ In 1963, Meyers et al., applied magnetic targeting technology to IONPs for targeted drug delivery.¹¹⁸ Zhang et al. further modified IONPs with folate on their surface and loaded them with cisplatin and si-GPX4. Folate-modified porous IONPs achieved ferroptosis and apoptosis in glioblastoma cells. This combination of gene therapy and chemotherapy effectively inhibited local glioma recurrence after surgery.¹¹⁹ In addition, IONPs possess different magnetic properties than bulk magnetite. Below 100 nm, these nanoparticles exhibit single-domain magnetism and have maximum coercivity compared to multi-domain bulk magnetite. Moreover, a further reduction in the size of the ions reduces their magnetic anisotropy energy. When the thermal energy matches the anisotropic energy, the magnetic moments randomly flip, resulting in the superparamagnetism of the nanoparticles. These superparamagnetic IONPs can be heated in an alternating electric field for applications in cancer targeting and thermal ablation.

Gadolinium chelate is the main clinical MRI T1 contrast agent. Gadolinium chelate releases free gadolinium ions in the body, which is toxic. In addition, the vast majority of gadolinium chelates used in clinical practice are not efficient in terms of cell labeling and cell stability. Gadolinium oxide (Gd₂O₃) has the characteristics of high longitudinal relaxation time relative ratio (r1), low cost, and easy synthesis. Ultra-small diameter Gd₂O₃ nanoparticles can concentrate a large number of magnetic ions in a small volume due to their high surface-to-volume ratio, thus providing a high signal-to-noise ratio.⁵⁹ The stability of the contrast agent determines whether it will decompose in a short period of time, affecting the imaging efficiency. The extremely small Gd₂O₃ nanoparticles synthesized with polymaleic acid as a stabilizer showed an excellent high r1 value, very small particle size, and excellent water dispersibility and stability.⁶⁰ Small-sized Gd₂O₃ can cross the blood–brain barrier. Small size Gd₂O₃ nanoparticles stabilized by polyacrylic acid were used to synthesize composite nanoparticles with a good relaxation effect, which could realize the integrated diagnosis and treatment of glioblastoma.⁶¹ Gd₂O₃ can be used as a sensitizer for radiotherapy. Gd₂O₃ promotes hydroxyl radical production in x-ray-irradiated aqueous solutions, increases intracellular reactive oxygen species (ROS) production, and activates the non-small cell lung cancer pro-death autophagy pathway.¹²⁰ Gd₂O₃ nanoparticles are highly photostable under the doping of luminescent rare earth ions. Modifying bombesin on the surface of Gd₂O₃ can target the gastrin-releasing peptide receptors highly expressed on the surface of prostate cancer cells and enhance the cancer targeting of Gd₂O₃. Then, in combination with the optical characteristics of Gd₂O₃, MRI/optical dual-modal imaging can be realized.¹²¹ 

Manganese dioxide (MnO₂) nanoparticles have good biocompatibility, and the Mn²⁺ produced by decomposition is soluble in water and easy to excrete and has low toxicity.⁶² MnO₂ nanoparticles can respond to a slightly acidic cancer microenvironment and can be used for stimulatory response release for targeted delivery. In addition, MnO₂ nanoparticles have shown unique advantages in regulating the cancer microenvironment. MnO₂ nanoparticles can react with reducibility GSH in cancer cells, resulting in oxidative stress. MnO₂ nanoparticles have catalase activity and can chemically react with H₂O₂ to generate O₂ and improve the hypoxic state of the cancer microenvironment. After the decomposition of MnO₂ nanoparticles, Mn²⁺ can undergo a Fenton reaction in the weak acid environment of cancers, catalyzing the production of hydroxyl radicals by H₂O₂, which leads to the death of cancer cells.⁶³ MnO₂ nanoparticles can increase the current response of biosensors, reduce the detection limit, and greatly improve the sensitivity of detection. MnO₂ nanomaterials have a strong light absorption capacity in the entire visible range and are widely used as fluorescence quenchers in fluorescent biosensors.⁶⁴ In addition, Mn²⁺ ions are used as contrast agents for MRI. Zhao et al. utilized MnO₂ nanoparticles as DNA nanocarriers, fluorescent quenchers, and intracellular glutathione (GSH)-activated MRI contrast agents. A dual-activated fluorescence/MRI dual-mode imaging platform was designed for cancer cell imaging.⁶⁵ 

Calcium-based carriers are considered one of the safest nanomaterials for use in cancer treatment. Their decomposition products are harmless components in the bloodstream. Most calcium-based functional materials exhibit reactivity within the TME and can interact with the cancer micro-acid environment, making them ideal for designing pH-responsive therapeutic systems.^110,122,123 CaP maintains its mineral structure at physiological pH and can dissolve within intracellular endosomes and lysosomes, enabling controlled drug delivery within cells. Upon dissolution, the calcium ions formed prevent particle accumulation and induce intracellular drug release. Qiu et al. utilized an in situ mineralization method to construct DOX-loaded CaP nanoparticles and further modified the surface with the Arg-Gory-Asp sequence. This modification allowed for efficient targeting of integrin αvβ3, which is highly expressed in ovarian cancers, leading to effective treatment of ovarian cancer [Fig. 4(a)].⁶⁹ By leveraging the acid-responsive decomposition property of CaCO₃ nanoparticles, responsive release and deep delivery of drug molecules in cancers can be achieved. For example, Zhu et al. developed pH-responsive nanoparticles (DNCaNPs) using CaCO₃ for efficient encapsulation of DOX and alkylated NLG919. DNCaNPs induce ICD in cancer cells under acidic conditions while inhibiting IDO1, thereby reducing the production of immunosuppressive molecules [Fig. 4(b)].¹²⁴ 

Upconversion nanoparticles (UCNPs) are excited by low-energy photons in the NIR range and emit high-energy photons in the UV or visible light range through an anti-Stokes process. Thus, UCNPs are excited by both long-wavelength and short-wavelength light emissions, thereby exhibiting great potential for use in many biological applications such as fluorescence microscopy, deep tissue bioimaging, nanomedicines, optogenetics, safety labeling, and volume display.^71–73,125 Song et al. constructed UCNPs with mitochondrial targeting by covalently modifying siRNAs to the surface of UCNPs using reactive oxygen-sensitive chemical bonds. Subsequently, mitochondrial targeting gene-encoded photosensitizers were constructed and transfected into cancer cells. After transfection and incubation, both UCNPs and gene-encoded photosensitizers accumulated in the mitochondria of cancer cells. Under NIR light irradiation, the emitted light from UCNPs stimulates the expression of protein photosensitizers, thereby producing ROS. Consequently, the controlled release of siRNA is triggered, enabling PDT and cascade gene therapy.⁷⁴ The deep tissue penetration capability of NIR light allows UCNPs to be excited and emit luminescence for in vivo imaging. By converting NIR light to UV light, UV light-based optical control in conjunction with UCNPs enables therapeutic strategies in vivo.

The development of nanomaterials in recent years has brought opportunities for the construction of practical biosensors to achieve rapid, sensitive, and accurate detection of biomarkers.^126,127 The electrode interface of biosensor based on inorganic nanomaterials has a high specific surface area, which can fix more active substances such as antibodies and aptamers, thereby improving the performance of biosensor. Inorganic nanomaterials can also take advantage of their good biocompatibility and electrical conductivity so that they can react with more targets and enhance the reaction process. Some inorganic nanomaterials can play a role in signal amplification due to their optical and magnetic properties. Biosensors constructed using the synergistic effect of nanocomposites tend to exhibit more stable and sensitive properties, further enhancing the detection signal and improving the confidence of the detection.¹²⁸ All kinds of inorganic nanomaterials, for example, AuNPs, CNTs, graphene, and IONPs, such as silica nanoparticles, have been widely used to load a variety of high levels of markers as a signal for immunoassay tracer, significantly improve the individual raw material identification of the combination of electrochemical markers, and eventually improve the sensitivity of immune analysis.^129–133 Biosensors based on novel functionalized nanomaterials have become a research hotspot in the biomedical field. At present, the biosensors used for cancer biomarker detection mainly include electrochemical biosensors, optical biosensors, photoelectrochemical biosensors, piezoelectric biosensors, and aptamer sensors Table II.¹³⁴ 

The key to the construction of electrochemical biosensors is the fixation of biomacromolecules, which aims to make biomacromolecules in contact with the electrode surface and avoid the biomacromolecules falling off from the electrode surface. At present, many kinds of nanometer materials are used to implement biological macromolecular structures fixed on the surface of the electrode, effectively improving the performance of the sensor. Gold nanomodified CNTs are loaded with a high proportion of glucose oxidase and signaling antibodies as markers. After the immune response of antigen and antibody, the glucose substrate can be catalyzed and the sensitive detection of cancer markers can be realized.¹³⁵ Modification of electrodes is the most promising approach to improve the sensitivity, selectivity, analyte adhesion, detection limit, and dynamic range of electrochemical sensing devices. Graphene has good electrical conductivity and a large specific surface area loaded with a large number of QDs. QD–graphene composite nanomaterials as electrochemical markers can amplify the sensor signal to improve the sensitivity of the sensor.¹³² QDs would result in a significant increase in the effective surface area of the modified electrode to load biomolecules, thereby amplifying the resulting electrochemical signal.^151,152 Heidari et al.¹⁵³ immobilized CdS QDs on a glassy carbon electrode and introduced AuNPs into the process by forming a sandwich-type immune complex between p53 antigen and antibody to develop an efficient electrochemical immunosensor for ultrasensitive determination of p53 biomarkers. AuNPs can enhance the emission of CdS QD sensor in the presence of H₂O₂.

The sensitivity of optical biosensor detection can be optimized by enhancing the conversion and amplification of optical signals. Many inorganic nanomaterials have stable and unique optical properties, which make them widely used in fluorescent sensors.^154,155 For example, AuNPs, silver nanocluster and graphene, CNTS, and QDs have been widely used to construct optical biosensors to improve the sensitivity of target detection.¹⁵⁶ Graphene has excellent fluorescence quenching. When the distance between the donor fluorophore and graphene is less than 10 nm, the graphene will absorb the excitation fluorescence of the fluorophore under the excitation light, thereby quenching its fluorescence. However, once the fluorophore is separated from the graphene, the quenching effect disappears.^157,158 The study of Maillard et al. found that metal–organic framework and graphene can interact with fluorescent substances for fluorescence quenching and can be used to construct fluorescent biosensors.¹⁵⁹ Williams et al. developed a fluorescent nano-sensor for the cancer biomarker urokinase plasminogen activator using single-wall CNTs. The limit of detection was 100 pM.¹⁶⁰ Kalkal et al. used amine-functionalized and nitrogen-doped graphene QDs as energy donors and AuNPs as energy acceptors to detect neuron-specific enolase. Under optimal conditions, the fluorescent biosensor has a response time of 16 min, a linear detection range of 0.1 pg/ml–1000 ng/ml, and an extremely low detection limit of 0.09 pg/ml.¹³⁶ 

In contrast to conventional electrochemical and optical detection techniques, photoelectrochemical biosensors employ light as the excitation source and electrical signals as the output detection signal. Due to the different and completely separated energy forms between the excitation source (light) and the output signal (photocurrent), the photoelectrochemical biosensor can effectively reduce the interference of background signals and improve the detection sensitivity.^161,162 In addition, compared with spectral detection techniques, photoelectrochemical biosensors do not require complex and expensive equipment and are characterized by low cost and easy miniaturization. Photoelectric conversion unit and biological recognition unit are two important components of photoelectric chemical biosensors. The photoelectric conversion unit is mainly a photoactive material, which converts biological or chemical information into observable photoelectric signals. The selection of photoactive materials is the core part of the construction of photoelectrochemical biosensors, which directly affects the analytical performance of the sensors.¹⁶³ Among various nanomaterials, inorganic nanomaterials have been widely used in the construction of photoelectrochemical biosensors due to their advantages of high fluorescence quantum yield, simple preparation, and easy control of size morphology. They mainly include metal oxides and semiconductor QDs.^164,165 AuNPs and CdS QDs/TiO₂ can produce an energy resonance transfer effect, resulting in the reduction of photocurrent signal. When the target prostate specific antigen (PSA) is introduced, the specific recognition between the PSA and the aptamer leads to the detachment of the AuNPs/Gn-labeled PSA aptamer from the electrode surface, and the photocurrent signal of the system is recovered, so as to achieve the purpose of PSA detection.¹³⁷ Dong et al. used CdSe-sensitized TiO₂ as a photochemical substrate and AuNPs-Dpa-melanin CNSs modified with AuNPs as a signal quench agent. Based on the specific recognition between PSA antibody and PSA labeled with AuNPs-Dpa-melanin CNSs, a competitive immune-sensing platform was designed for the detection of PSA. The immunosensor has a wide linear range with a detection limit of 2.7 pg/ml.¹⁶⁶ 

Piezoelectric biosensor is a novel bioanalysis method that combines highly sensitive piezoelectric sensors with specific biological responses. The electromechanical conversion properties of piezoelectric materials can convert external stimuli (ultrasound, pressure, motion) into electrical energy and can sensitively sense environmental changes for real-time biosensing.¹⁶⁷ The most commonly used piezoelectric crystal is quartz crystal. Quartz crystal has the advantages of high sensitivity and simple fabrication, which responds not only to the quality change of the crystal surface but also to the mechanical energy change of the quartz crystal surface. The lower the number of graphene layers, the better the sensitivity of the sensor. Graphene sheets were fabricated by chemical vapor deposition and then transferred to a 36° Y–90° X quartz substrate. Immobilization of aptamers can be achieved by cross-linking chitosan with glutaraldehyde. The detection limit of this biosensor for lipopolysaccharide was 3.53 ng/ml.¹³⁸ Yagati et al. used reduced graphene oxide (RGO) and AuNPs to detect the cancer necrosis factor on indium tin oxide micro-disk electrodes, achieving a detection limit of 0.78 pg/ml and a sensitivity range between 1 and 1000 pg/ml.¹⁶⁸ 

At present, there are many shortcomings in traditional medical imaging technologies such as x ray and CT. The resolution and clarity may not meet the medical needs, which may lead to the missed or misdiagnosis of some minor lesions. In addition, traditional medical imaging technology mainly detects macroscopic lesions after the occurrence of diseases, which is difficult to detect and diagnose at the early stage of diseases or at the molecular and cellular levels, which affects the early diagnosis of diseases. The application of nanotechnology in medical imaging has significantly improved the sensitivity of early cancer detection and diagnosis. Inorganic nanoparticles can be used as contrast agents to improve image resolution in CT, MRI, nuclear medicine imaging, and other technologies.¹¹⁵ Due to their small size and high stability, these nano-contrast agents circulate more smoothly in the body and are less prone to aggregation, resulting in sharper, higher-resolution images. In addition, FLI, PAI, and various combined imaging modes have been further developed by taking advantage of the characteristics of inorganic nanometers (Table II).

CT imaging has the advantages of high sensitivity, non-invasiveness, and depth-independent imaging. However, the currently commonly used iodine contrast agents have limitations in defining cancer margins accurately during surgery and radiotherapy due to their rapid clearance. Consequently, the conversion of real-time CT diagnosis results into precise therapeutic interventions has become a pressing clinical challenge. A large number of inorganic metal nanomaterials can be ideally used as CT contrast agents due to their unique physical, chemical, and biological properties, for example, gold-based, bismuth-based, lanthanide-based, and transition metal-based inorganic nanomaterials.¹⁶⁹ AuNPs exhibit strong light scattering and absorption properties, which can be adjusted by controlling their size and internal structure. This adjustability makes it relatively straightforward to tailor the properties of AuNPs for specific imaging modes. Moreover, AuNPs possess a high x-ray absorption coefficient, making them suitable for CT imaging.^170,171 PEG-modified gold nanorods (GNRs) can be used for in vivo CT imaging for enhancing targeting and improving guidance in surgical irradiation.¹⁷² The ability of CT to specifically recognize cancer calcification aids in monitoring the therapeutic effect in medical imaging. The development of calcium-based nanomaterials not only serves as effective therapeutics but also accelerates cancer calcification, enabling visual monitoring of imaging features. Upon administration of calcium-based nanomaterial drugs, a noticeable enhancement in CT signal and brightening of the cancer area can be observed. Over time, as calcification intensifies, the CT signal further increases [Fig. 5(a)].¹⁷³ The use of inorganic nanomaterials for CT contrast agents can be combined with PPT, PDT, chemotherapy, radiotherapy, gas therapy, etc., to achieve CT image-guided cancer treatment.^169,174 Gao et al. designed a chemo-photothermal synergistic therapy based on AuNRs@MIL-101 nanostructure by combining AuNRs with MIL-101 metal–organic frameworks and then mounting carboxylatopillar[5]arene-based supramolecular gates. The AuNRs inside endow the as-synthesized nanoplatform with high photothermal efficiency under the 808 nm laser irradiation and desirable CT imaging.¹⁷⁵ Cheng et al. linked zinc protoporphyrin IX (ZP) to Bi₂S₃ nanorods through a thermo-responsive polymer to form BPZP nanosystems. Under the irradiation of NIR laser, the heat released by the Bi₂S₃ nanorods in the BPZP nanosystem can be used for PTT, and combined with the CT imaging performance of Bi₂S₃ nanorods, an integrated diagnosis and treatment platform can be realized.¹⁷⁶ However, efforts are still needed to use inorganic nanomaterials for CT imaging, such as the development of endogenous biomarker-responsive CT contrast agents to improve their sensitivity and specificity, biocompatibility, and toxicity in vivo.¹⁶⁹ 

MRI is a non-invasive, non-ionizing imaging method used to obtain physiological and pathological information about living tissues by measuring proton relaxation rates in water. The commonly used contrast agents in MRI are T1 contrast agents (gadolinium-based and manganese-based materials) and T2 contrast agents (iron-based materials). Among them, superparamagnetic IONPs are widely employed as excellent MRI contrast agents for diagnosing related diseases. These IONPs exhibit superparamagnetism and serve as drug carriers and nuclear magnetic contrast agents. They offer enhanced specificity, a prolonged half-life in the body, and potential applications in cancer therapy.^177–179 T2 imaging of ferric oxide nanomaterials modified with different ligands at different time points can effectively detect changes in cancer uptake over time in vivo.¹⁸⁰ Liu et al. designed a straightforward and efficient CIZS@FeOOH nanoplatform using CuInZnS quantum dots (CIZS QDs) and FeOOH. FeOOH effectively quenches the fluorescence of nanomaterials, but its reduction triggered by GSH induces fluorescence recovery of CIZS QDs, activating the system. The overexpression of the NIR regions of GSH and CIZS QDs in cancer cells enables the probe to perform specific in situ imaging of cancer sites. In addition, the Cu-based peroxidase catalytic performance and Fe²⁺-triggered Fenton reaction simulate the generation of OH from H₂O₂, enabling efficient CDT. Furthermore, the Fe³⁺ Fenton reaction can be used as a T1-weighted MRI agent, making it a valuable tool for therapeutic diagnosis [Fig. 5(b)].¹⁸¹ 

Nuclear medicine imaging requires suitable radiotracers for imaging functionality. Due to the low spatial resolution of nuclear medicine imaging, it is difficult to pinpoint the anatomical location of the radiotracer with nuclear medicine imaging alone. Labeling functional inorganic nanoparticles with medical radionuclides can realize the functional combination of nuclear medicine imaging and other imaging methods. For example, radiolabeled ultra-small Fe₃O₄ nanoparticles targeting the integrin avβ3 receptor can achieve MRI/SPECT multi-functional imaging.¹⁸² PET/NIRF multi-functional imaging can be achieved by mixing ⁶⁴Cu with ferritin subunits labeled on the surface with the targeting ligand RGD4C and the fluorescent motif Cy5.5, respectively.¹⁸³ A PET/FLI/CL/CRET multimodal imaging nanoprobe was developed by assembling copper sulfide on the surface of HMSN nanoclusters labeled with Zr and supported porphyrin molecules.¹⁸⁴ BizSes nanosheets with FeSe₂ nanoparticles modified with FeSe₂ nanoparticles labeled with ⁶Cu can be used for PET/MR/CT/PA four-modal imaging in cancer-bearing mice.¹⁸⁵ The use of inorganic nanometers to combine with radionuclides can improve the efficiency of imaging.¹⁸⁶ The use of low-generation dendritic macromolecules to embed AuNPs and label ⁹⁹Tc after FA functionalization on the surface can be used for SPECT/CT targeted bimodal imaging of cancers overexpressing FA receptor, with high targeting.¹⁸⁷ The use of inorganic nanometers to realize the adjustable blood circulation time of nuclear medicine imaging can obtain better cancer selectivity and detection time window.¹⁸⁸ In order to slow down the uptake of Fe₃O₄ nanoparticles by the liver and prolong the blood circulation time, Gao et al. designed a PEG ligand containing an RGD peptide linked by disulfide bonds and a “self-peptide” sequence to modify small-sized Fe₃O₄ nanoparticles and label the radionuclide ⁹mTc to construct a cancer microenvironment-responsive SPECT/MRI dual-modal probe based on GSH-triggered nanoparticle aggregation for cancer enhancement imaging. Auto-peptides can help nanoprobes avoid RES uptake before they reach cancer lesions in the blood circulation.¹⁴¹ 

FLI is a powerful molecular imaging technique in which specific probes (i.e., fluorophores) are emitted at lower excitation energies. Traditional FLI has limited resolution and can only penetrate superficial tissue up to 2–3 mm due to the limitations of light. To enable FLI for broader clinical applications, fluorophores must possess properties such as high fluorescence quantum yield, photostability, and resistance to degradation in biological systems. Therefore, the development of fluorophores with these properties is necessary for cancer imaging.^189,190 QDs offer advantages over organic dyes and fluorescent proteins as they provide a wider range of emission spectra covering both visible and NIR wavelengths, have greater absorption coefficients, and exhibit a much higher photostability. Therefore, QDs have been extensively investigated as contrast agents for FLI.^191,192 By connecting fluorescence or QDs, MSNs can serve as contrast agents for in vivo imaging. Benezra et al. developed an MSN probe encapsulating optical dyes and radioactive iodine oil, enabling direct targeting of human melanoma. This approach allows real-time monitoring of MSN clearance, lymphatic drainage patterns, and lymph node metastasis.¹⁹³ Lee et al. synthesized MCM-41 nanoparticles ranging from 50 to 100 nm and modified them with indocyanine green (ICG). Injecting labeled ICG-MSNs into mice resulted in strong fluorescence signals detected in the liver after 3 h. Since the excitation and emission wavelengths of ICG are 800–820 nm, ICG-MSN imaging results in a lower image signal-to-noise ratio and higher tissue penetration.¹⁴² Combining iron-gold nanoclusters with glucose oxidase enables targeted imaging of cancers and ferroptosis [Fig. 5(c)].¹⁹⁴ 

PAI is an innovative non-invasive imaging technique that leverages the photoacoustic effect to overcome the challenges posed by high photon scattering in biological tissues. This enables higher spatial resolution in the resulting images. When a photothermal conversion reagent absorbs light, it generates transient thermoelastic expansion, emitting ultrasound and creating a photoacoustic image.^195,196 PAI technology also facilitates real-time monitoring and surgical guidance. In most cases, a photothermal conversion reagent can be used for both PAI and PTT, integrating disease diagnosis and treatment, particularly for conditions such as cancers.¹⁹⁷ The photoacoustic signal is proportional not only to the light absorption of the nanoparticle solution, but also to the surface volume ratio of the nanoparticle. GNRs, which are 5–11 times smaller than the conventional size, exhibit 3 times more stability and 3.5 times the PAI signal intensity.¹⁴³ Li et al. developed Janus gold nanorod platinum (JAuNR-Pt) nanomotors driven by nano-hydrogen peroxide to enhance photoacoustic imaging in the NIR region for deep tissue treatment and effective cancer management. The self-promotion of JAuNR-Pt nanomotors enhances cellular uptake, accelerates lysosomal escape, and promotes the continuous release of cytotoxic Pt²⁺ into the nucleus, leading to DNA damage and cell apoptosis [Fig. 5(d)].¹⁹⁸ 

To avoid potential false positive interference and improve accuracy in cancer location imaging, nanoprobes can be used in combination, leveraging multiple imaging modes. MRI is a powerful non-ionizing radiation technique with high spatial resolution, optimal soft tissue contrast, and no depth limitations. However, it has limited sensitivity and longer imaging times. FLI provides high-intensity real-time images, but its penetration in deep tissues is poor. As a rapidly developing imaging method, PAI offers deep tissue penetration and high optical contrast but lacks suitable developers for broader applications. Integrating two or more imaging technologies onto a single nanoprobe can provide diagnostic images with high spatial resolution, contrast, and sensitivity for cancer treatment. For example, the combination of Fe₃O₄ and AuNPs can enable CT/MRI dual-modal imaging diagnosis.¹⁹⁹ Tang et al. developed a pH-activated nanoprobe (GNPs-CKL-FA) using a NIR fluorophore (Cy5.5), a pH-sensitive keto linker, and AuNPs modified with FA (GNP). GNPs-CKL-FA selectively binds to cancer cells through folate-mediated internalization, enabling targeted cancer CT imaging. In the acidic microenvironment of the cancer, the keto linker of GNPs-CKL-FA is hydrolyzed. This releases Cy5.5, thereby emitting red fluorescence and generating a fluorescent signal [Fig. 6(a)].²⁰⁰ AIE is a unique photophysical phenomenon where a fluorophore exhibits enhanced emission within an aggregate. Huang et al. used AIE and BP to construct a novel multi-functional therapeutic diagnostic nanoplatform (BP@PEG-TTPy). Based on in vitro and in vivo experiments, BP@PEG-TTPy exhibited an excellent performance in cancer therapeutic diagnostics, involving NIR FLI-PTI dual image-guided synergistic PDT-PTT phototherapy [Fig. 6(b)].²⁰¹ Moreover, the integration of multiple imaging modalities and therapies provides a new approach to cancer treatment. Sun et al. developed a multi-functional diagnostic and therapeutic nanoplatform using AuNR dimers and UCNPs that combines PTT and PDT, as well as MRI, PAI, and CT imaging. This platform exhibits a significant remarkable cancer-targeting capability through the EPR effect and achieves efficient cancer treatment in vivo [Fig. 6(c)].²⁰² 

The biggest challenge of radiotherapy is how to efficiently deposit the energy of high-energy rays at the cancer site, so as to achieve sensitization of radiotherapy while reducing the damage of ionizing radiation to surrounding normal tissues. Metal–inorganic nanomaterials can promote the deposition of radiation energy at cancer sites and improve the efficiency of radiotherapy.^203,204 The use of AuNPs in radiotherapy causes an increase in mutations in plasmid DNA caused by x-ray irradiation.²⁰⁵ Synergistic radio-chemotherapy was achieved by combining the radio-sensitizing effect of Au nanorods and the apparent anti-cancer activity of Se nanoparticles.²⁰⁶ The combination of targeted modified inorganic nanomaterials and radioactive isotope elements can achieve precise delivery of radioisotopes at cancer sites and reduce the radiotoxicity of other normal organs.²⁰⁷ The use of nanomaterials to deliver biomarkers and then combine them with radioactive isotope elements enables individualized radiotherapy.²⁰⁸ 131I is added to CuS nanoparticles and modified with PEG. CuS/[131I]PEG nanoparticles target migration into sentinel lymph nodes, enabling not only radiotherapy and PTT, but also effective inhibition of cancer metastasis (Table II).¹⁴⁴ In addition, nanomaterials with cancer microenvironment modulation functions are designed to overcome hypoxia-related radiotherapy tolerance and improve radiotherapy efficacy.²⁰⁹ Au@MnO₂–PEG nanoparticles also use the MnO₂ shell to induce the decomposition of endogenous H₂O₂ in the cancer microenvironment to generate oxygen, which is a highly effective radiosensitizer for radiotherapy of hypoxic cancers.²¹⁰ 

Conventional anti-cancer chemicals have certain limitations, such as strong hydrophobicity, significant side effects, lack of targeting, and potential adverse reactions, including bone marrow transplantation. These limitations greatly limit the practical use of chemotherapy drugs. The rapid progress of nanotechnology has provided new strategies for the delivery of chemotherapy drugs.^211,212 Nanodrug carriers can actively target cancer cells by modifying selective ligands on the surface of nanoparticles or through passive targeting mediated by the EPR effect. This enhances the specific selectivity of drugs toward cancer cells, increases drug concentration in the target area, reduces distribution in non-targeted sites, and minimizes adverse reactions.²¹³ Acute kidney injury (AKI) is a common and fatal side effect of ROS-associated cancer chemotherapy. Weng et al. designed pH-dependent cerium dioxide nanoparticles (CNPs) that prevent chemotherapy-induced AKI without interfering with the effects of chemotherapy. In the neutral pH environment of normal cells, CNPs decompose H₂O₂ absorbed on the surface and re-expose the active catalytic site of CNPs to restore the ROS balance, thereby protecting the renal tubules and reducing the toxic side effects of chemotherapy drugs. However, in the acidic environment of cancer cells, H⁺ destroys the active catalytic site of CNPs, limiting their antioxidant effect and preventing ROS scavenging. This does not compromise the cancer-killing ability of chemotherapeutic drugs [Fig. 7(a)].²¹⁴ 

Drug resistance in cancer cells is a major factor contributing to the failure of chemotherapy and can lead to cancer recurrence and progression. The use of nanomaterials enables targeted damage to drug-resistant cancer cells. Wang et al. constructed an amorphous CaP-based nanoplatform that targets ferroptosis and apoptosis in drug-resistant cancer cells by blocking the McT4-mediated lactate efflux [Fig. 7(b)].²¹⁵ Inorganic nanoparticles, such as HMSNs, have been widely employed as drug carriers due to their low toxicity, high biocompatibility, and high drug-loading capacity.²¹⁶ Zhang et al. modified the surface of MSNs with HA through disulfide bonds, creating a CD44-targeted drug delivery system. Since cancer cells overexpress CD44 receptors, HA-functionalized MSNs can specifically recognize and internalize drugs in cancer tissues, enhancing the anti-cancer effect.¹⁴⁵ Magnetic inorganic nanomaterials can serve as effective drug delivery carriers due to their magnetic response properties. With the assistance of an external magnetic field, targeted drug delivery and controlled release can be achieved, significantly improving delivery efficiency and therapeutic outcomes.²¹⁷ Jon et al. synthesized PEG cross-linked SPIONs for electrostatic adsorption of chemotherapy drugs. Its anti-cancer effect was approximately eight times higher than that of simple chemotherapy drugs. In addition, inorganic nanoparticles can carry multiple drugs simultaneously, reducing cancer resistance to a single drug and diversifying treatment options, thereby increasing the efficiency and therapeutic effect of chemotherapy.²¹⁸ Nosrati et al. modified Au nanoparticles on the surface of Bi₂S₃@BSA nanoparticles to produce Bi₂S₃@BSA-Au. Bi₂S₃@BSA-Au can enhance the contrast of CT images and act as a radiosensitizer. Under x-ray irradiation, Bi₂S₃@BSA-Au can increase the production of ROS, damage DNA, and enhance the therapeutic effects of methotrexate (MTX) by inhibiting dihydrofolate reductase. Moreover, curcumin (CUR) can be used alongside chemotherapeutic agents to enhance chemotherapy and protect normal tissues from radiation damage. The simultaneous encapsulation of MTX and CUR with Bi₂S₃@BSA-Au nanoparticles resulted in the formation of Bi₂S₃@BSA-Au-BSA-MTX-CUR nanoparticles. This multi-functional nanosystem serves as a contrast agent, drug carrier, and radiation sensitizer, offering a safe, practical, and effective approach to cancer treatment [Fig. 7(c)].²¹⁹ 

The success of gene therapy depends heavily on the transfection ability of gene delivery vectors. Successful gene therapy requires the safe, efficient, and selective delivery of target genes to the lesion site, ensuring their sustained and efficient expression.²²⁰ Gene delivery strategies are generally categorized into viral and non-viral strategies. The development of viral vectors has been hindered by limitations including low safety, limited gene load, and immunogenicity. In recent years, non-viral vectors, such as cationic liposomes, PEI, high molecular weight polymers, and inorganic nanomaterials, have gained attention due to their superior safety, high gene loading capacity, and ease of production scalability. Inorganic nanoparticles are emerging as potential carriers for nucleic acid delivery, offering advantages over traditional lipid and viral carriers, such as adjustable size, stable storage, targeting capability, and the ability to cross biological barriers, with favorable pharmacokinetic and toxicity profiles.²²¹ Xu et al. synthesized large amino acid-mimicking guanidine-functionalized CDs (LAAM GUA-CDs) using arginine and dopamine hydrochloride as precursors. LAAM GUA-CDs effectively loaded siRNA through multiple hydrogen bonds between guanidine and phosphate groups in siRNA. By exploiting the interaction between amino acid groups and overexpressed LAT1 on cancer cells, LAAM GUA-CDs/siBcl-2 specifically targeted the cancer site. The results showed that LAAM GUA-CDs/siBcl-2 achieved 82% knockdown of Bcl-2 protein expression in cancer cells and inhibited cancer progression by 68%. The transfection efficiency of LAAM GuA-CDS/siBcl-2 was twice that of commercial Lipofectamine 2000, with no significant toxicity observed [Fig. 8(a)].¹⁴⁶ AuNPs and magnetic nanoparticles can easily be modified with macromolecules or polymers, serving as carriers for gene therapy and facilitating the integration of multiple functions. These carrier platforms can deliver various gene therapeutic agents to the cancer site, effectively killing the cancer and fulfilling the role of cancer treatment.^222–224 Peng et al. developed Fe₃O₄/PEI-PEG NGs by combining ultra-small IONPs (Fe₃O₄NPs) with PEI (800 Da) and PEG (400 Da)-diacrylate. These NGs were then loaded with transforming growth factor-β1 (TGF-β1) siRNA for gene therapy. The Fe₃O₄/PEI-PEG NGs not only served as an in vivo MR imaging tool but also effectively suppressed the growth, invasion, and lung metastasis of sarcoma by silencing the TGF-β1 gene [Fig. 8(b)].²²⁵ Moreover, leveraging the properties of inorganic nanoparticles to construct a nano-diagnosis and treatment platform with imaging capabilities allows for visualization of the gene therapy process and potential parallel diagnostic capabilities. For example, Liu et al. utilized polycationic PGEA (ethanolamine-functionalized polyglycidyl methacrylate) and functionalized dextran QDs to construct a nanohybrid (DQ-PGEA) as a safe and efficient gene carrier. The efficacy of this prepared gene delivery system was verified using the anti-cancer gene p53 in a mouse model of breast cancer. Simultaneously, QDs enabled FLI, enabling real-time tracking of the gene delivery process, visualization of the treatment process, and FLI-guided gene therapy [Fig. 8(c)].²²⁶ 

Immunotherapy, a revolutionary approach to cancer treatment, aims to activate the innate and adaptive immune system to eliminate cancers.²²⁷ It involves various mechanisms, including reducing cancer-related antigen display, overexpressing immune checkpoint molecules, and recruiting and inducing immunosuppressive cells. Different immunotherapies have been developed to target specific stages of the cancer immune cycle, such as ICB therapy, chimeric antigen receptor T cell (CAR-T) therapy, cytokine-based immunotherapy, and cancer vaccine-based immunotherapy.^228–230 Although these treatments have shown promising results in clinical trials, challenges remain, such as limited sustained response rates, immune-related side effects, and abnormal clinical reactions. With the development of nanotechnology, inorganic nanoparticles have received widespread attention in cancer immunotherapy due to their advantages in targeted drug delivery, precise targeting of drug delivery, simple surface functionalization, and biological activity.²³¹ 

ICB therapy has revolutionized cancer treatment by targeting the T cell immunosuppressive pathway. By blocking the interaction between immune checkpoints on cancer cells and immune cells, this therapy eliminates the inhibitory effect of cancers on immune cells. This treatment method restores the monitoring and killing ability of immune cells against pathological cancers by applying immune checkpoint pathway inhibitors, such as antibodies against programmed cell death protein 1 (PD-1), programmed cell death ligand 1, and cytotoxic T lymphocyte-associated protein 4.²³² While these inhibitors have shown efficacy in various cancers, they also present certain challenges. Nanotechnology offers a solution to overcome the limitations associated with immune checkpoint inhibitors, providing targeted delivery, reduced side effects, and enhanced efficacy.²³³ 

In the TME, low immunogenicity and immunosuppression can hinder the effectiveness of ICB treatment. To address this, Zhao et al. designed a manganese-catalyzed immunotherapy that activates the TME, working synergistically with ICB therapy to effectively eradicate cancers. They used cancer cell membranes (CMs) to encapsulate manganese oxide nanoparticles with multi-enzyme mimicking capabilities. These CM@Mn nanoenzymes exhibit oxidase activity in acidic environments, generating toxic hydroxyl and superoxide free radicals to induce cancer cell death and ICD. Moreover, the release of Mn²⁺ in response to the TME directly promotes dendritic cell maturation and repolarization of macrophages toward the M1 phenotype, thereby transforming the immunosuppressive TME into an immune-activated environment. In addition, the catalase activity of the nanoparticles alleviates cancer hypoxia, contributing to the TME reversal process. Finally, in the presence of PD-1 checkpoint blockade, a cancer-specific T cell-mediated anti-cancer response is triggered, inhibiting the growth of primary and metastatic cancers and inducing long-term immune memory effects [Fig. 9(a)].²³⁴ 

Chimeric antigen receptor (CAR) T cell therapy has emerged as a successful immunotherapy approach following ICB therapy. CAR-T involves genetically engineering T cells to express antigen receptor proteins on their surface, enabling them to recognize and target specific cancer antigens, thus enhancing cellular immunotherapy.²³⁵ Despite its increasing attention, CAR-T immunotherapy is only beneficial for hematological malignancies. In addition, CAR-T immunotherapy has proven ineffective against solid cancers across many clinical cases. Similar to ICB therapy, the occurrence of these adverse effects can be attributed to challenges such as the low efficiency of CAR-T cell transport, their inability to persist effectively, and the presence of a complex immunosuppressive TME. Factors such as insufficient CAR-T cells and limited infiltration of CAR-T cells also contribute to these challenges.

Significant efforts have been devoted to improving CAR-T therapy and ICB to address these limitations. Zhou et al. have engineered CAR-T cells by splitting the chimeric antigen receptor functional block into two modules and installing light control components into the two modules.¹⁴⁷ When exposed to blue light, the two modules combine to form a fully functional chimeric antigen receptor. Unlike traditional CAR-T cells that are activated upon binding to specific antigens, these engineered LiCAR-T cells can be precisely controlled by blue light, offering spatiotemporal control over their activation. Furthermore, researchers have successfully integrated UCNPs with LiCAR-T cells to achieve wireless manipulation using nano-photogenetic technology. UCNPs possess unique fluorescence characteristics, allowing them to convert NIR light into short-wavelength visible light through UCNPs. By using UCNPs as in situ light converters, NIR excitation can be converted to blue light, activating LiCAR-T cells. This method overcomes the limitations of traditional photogenetic tools that rely on visible light excitation with shallow tissue penetration, thus achieving wireless manipulation of nano-photogenetic technology. Moreover, this UCNP/LiCAR-approach effectively mitigates side effects associated with traditional CAR-T cell therapy, such as “on target/off cancer” cytotoxicity and cytokine release syndrome, significantly improving the safety of cell immunotherapy.

Cytokines, which are generally produced by stimulated immune cells, serve as molecular messengers that facilitate communication between immune system cells and coordination with target antigens. They possess regulatory and effector functions in various diseases, making cytokines and their receptors potential candidates for immunotherapy.²³⁶ In the clinical treatment of cancer, three primary types of cytokines are used: interleukins, interferons, and granulocyte–macrophage colony-stimulating factors. These cytokines can control cancer growth directly through antiproliferation and proapoptotic effects, or indirectly by activating the cytotoxic effect of immune cells on cancer cells.²³⁷ Cytokine therapy holds great potential for cancer immunotherapy, but its development is hindered by challenges such as the possibility of inducing a “cytokine storm” when systemically administered. At the same time, due to rapid enzymatic degradation and clearance, the half-life of cytokines in plasma is very short, and they have a strong targeting effect. To compensate for their short half-life, systemic administration of cytokines requires relatively large doses, which can lead to acute toxic side effects. Nanoparticles can deliver cytokines to specific tissues and continuously release them for hours or even days, with minimal adverse effects on healthy tissues.²³⁸ In the case of patients with sarcomas, certain cytokines such as cancer necrosis factor (TNF) show great potential as effective cancer therapies. However, high concentrations required to kill cancer cells often result in damage to healthy cells. Therefore, TNF is currently limited to the treatment of limb sarcomas and cannot be administered systemically. To address this issue, Curnis et al. developed a targeted anti-cancer cytokine therapy called NGR-TNF. This novel therapy NGR-TNF combines TNF with the Asn-Gly-Arg (NGR) peptide, a relatively short protein-coding fragment, that acts as a guide, directing drugs specifically into cancer blood vessels to induce cancer damage. Further research revealed that adsorbing AuNPs smaller than poliovirus onto NGR peptides could enhance the quick and efficient delivery of drugs to cancer blood vessels, leading to cancer damage. This novel super-NGR-TNF therapy effectively slows down cancer growth in cancer-prone mice without any significant side effects. This study also shows that the drug can safely and effectively reach cancer tissue and exert its therapeutic effects even at low levels.²³⁹ 

Nano-vaccines are a type of cancer vaccine utilizing nanomaterials as carriers to deliver specific antigens and adjuvants for the treatment or prevention of cancers.²⁴⁰ They have several advantages. The size of nanomaterials enables their concentration in lymph nodes, spleen, and lymphatic organs. In addition, their size allows for easy uptake by antigen-presenting cells (APCs) and targeted delivery of antigens and immune activators to specific immune cells involved in the “cancer immune cycle.” This process activates tumor-associated antigen (TAA)-specific T cells and stimulates the acquired immune response, leading to the killing of cancer cells.^241,242 Compared to traditional subcutaneous immunization with naked peptides, the pre-loading of antigen peptides and adjuvants onto nanoparticles before subcutaneous inoculation can significantly enhance the generation of new antigen-reactive T cells, leading to a nearly 30-fold increase. This approach effectively suppresses cancer growth. Inorganic nanomaterials, with their low biodegradability and stable structure, offer targeted accumulation in lymph nodes and efficient release in APCs due to their unique physical and chemical properties. These materials have wide applications in nano-vaccines. Meng et al. developed a nano-delivery platform that simultaneously delivers antigens and adjuvants to different intracellular compartments to enhance the immune response. Their nano-vaccine, known as IONP-C/O@LP, combines CpG, ovalbumin peptide (OVAP), and magnetic IONPs to form an IONP-C/O complex, which is then coated with a lipid membrane (L) containing a targeted dendritic cell cyclic peptide (P30). IONP not only endows nano-vaccines with detectability through MRI but also exhibits adjuvant effects through the production of intracellular ROS. Upon subcutaneous injection, IONP-C/O@LP effectively enters and activates dendritic cells, thereby stimulating local and systemic anti-cancer immune responses. This immune stimulation effectively inhibits cancer growth, positioning it as a potential therapeutic or preventive vaccine [Fig. 9(b)].²⁴³ Liu et al. inhibited B16-OVA cancer growth by complexing CDs with the model antigen ovalbumin. To enhance antigen binding, GCD was modified with maleimide caproic acid, resulting in GMal. Simultaneously, cancer cell-derived antigens (B16F10-Ag and CT26-Ag) were prepared using PTT to induce ICD. Subsequently, two nano-vaccines, namely GMal + B16F10-Ag and GMal + CT26-Ag, were prepared by binding GMal with either B16F10-Ag or CT26-Ag. These nano-vaccines displayed strong fluorescence, excellent cell permeability, and biocompatibility. Experiments showed that these nano-vaccines were internalized by dendritic cells, promoting their maturation. Moreover, they exhibited cancer-specific recognition of melanoma or colon cancers. When combined with ICB, these nano-vaccines significantly inhibited cancer growth and even eliminated cancers [Fig. 9(c)].²⁴⁴ 

The key component of PTT is the photothermal conversion agent, responsible for converting light energy into heat energy upon exposure to external light sources such as NIR light, thereby killing cancer cells. An excellent photothermal conversion reagent should be highly efficient in converting light into heat, exhibit strong absorption of NIR light with tissue-penetrating capacity, and achieve effective cancer enrichment.²⁴⁵ Nanoscale photothermal conversion reagents offer effective cancer enrichment through the EPR effect or active targeting of solid cancers, benefiting from the unique properties of nanomaterials. Moreover, their photothermal conversion efficiency is generally higher than that of small molecule photothermal conversion reagents.¹⁹⁵ Photothermal conversion agents formed from inorganic nanomaterials offer advantages such as high stability and strong photothermal conversion efficiency. For example, AuNPs have received widespread attention due to their ideal biocompatibility and strong LSPR effects. Carbon-based nanomaterials possess strong light absorption in the NIR region and exhibit remarkable stability, maintaining their light absorption performance even after prolonged irradiation. In 2005, Professor Dai Hongjie of Stanford University pioneered the application of single-walled CNTs (SWCNTs) in in vitro cancer PTT, achieving selective cancer-targeting through PEG modification and folate receptor coupling. In a subsequent study in 2009, the use of SWCNTs for in vivo cancer PTT was successfully demonstrated.^246–248 Robinson et al. have developed nanoscale-reduced GO sheets with high NIR light absorption and good biocompatibility. Through non-covalent functionalization using amphiphilic PEG-based polymer chains, these rGO sheets exhibit good dispersion and low biological toxicity and have been effectively employed in cancer PTT.¹⁴⁸ Metal-based inorganic nanoparticles possess relatively stable properties, resistance to bleaching and degradation by light, and exhibit broad absorption bands and high extinction coefficients in the NIR region.²⁴⁹ Wei et al. rapidly synthesized carbon spheres with high photothermal conversion efficiency via pyrolysis of pyrogallol formaldehyde spheres. These carbon spheres, calcined at 700 °C, demonstrate a photothermal conversion efficiency of 54.2%, surpassing previously reported carbon-based photothermal agents. Moreover, they exhibit excellent biocompatibility and effective in vitro and in vivo cancer cell ablation through photothermal effects [Fig. 10(a)].²⁵⁰ Other inorganic nanomaterials such as Cu sulfur compounds (CuS, Cu_2–xS) have also been widely used in cancer PTT due to their high photothermal conversion efficiency and NIR absorption.²⁵¹ 

The principle involves photosensitizers absorbing photon energy, stimulating the transfer of energy to surrounding oxygen molecules, and generating ROS, leading to cancer tissue necrosis and cancer cell death.^252,253 Most photosensitive molecules have poor water solubility and are easily metabolized, hindering their ability to achieve effective PDT at cancer sites. There is an urgent need to develop intelligent PDT systems for the precise regulation of photosensitizer activity and cancer specificity.^254,255 Loading photosensitive molecules onto or inside nanoparticles enhances their delivery while reducing distribution in normal tissues. Wang et al. used hydrophobic interactions to load Ce6 onto the surface of UCNPs. The UCNPs in this system are capable of absorbing long waves at 980 nm and emitting short waves at 660 nm, while Ce6 is specifically excited by a wavelength of 660 nm. This system enables PDT under long-wavelength excitation, extending its effectiveness to deep tissues.¹⁴⁹ Hsiao et al. created a core–shell structure of Fe₃O₄/SiO₂ functionalized with an iridium-based group for enhanced MRI and singlet oxygen production, thereby inducing cell apoptosis. The hypoxic microenvironment in most malignant cancers often counteracts the photodynamic effects of photosensitizers. Ensuring a continuous oxygen supply to cancer cells is crucial for the effectiveness of PDT.²⁵⁶ To address this issue, Qi et al. attached black scales to cyanobacteria. Cyanobacteria, single-celled prokaryotes with a long history, possess the ability to produce oxygen through photosynthesis, similar to green plants. By subjecting the cyanobacteria to laser irradiation, they were able to continuously generate oxygen. Subsequently, under illumination, the generated oxygen triggers the production of ROS, ultimately achieving the objective of cancer treatment [Fig. 10(b)].²⁵⁷ 

SDT is a novel cancer treatment that utilizes ultrasound to activate acoustic sensitizers in cancer cells, leading to the production of cytotoxic ROS and subsequent damage to the cancer cells. Ultrasound, widely used in clinical diagnosis and treatment, offers advantages such as high focus, deep tissue penetration, and minimal damage to normal tissues.²⁵⁸ In comparison with PDT, commonly used for superficial cancers, ultrasound demonstrates notable benefits for treating deep-seated cancers. However, SDT still faces limitations, particularly the lack of an optimal sound-sensitizing agent. Traditional organic sound sensitizers have limitations such as poor tissue accumulation, low stability, limited bioavailability, and rapid clearance from the body, hindering their clinical translation. The application of nanomaterials presents an effective means to enhance SDT efficiency and potentially overcome the limitations of conventional approaches.^259,260 Zhang et al. designed hollow Fe₃O₄ nanoparticles as carriers for delivering hematoporphyrin, a sound-sensitive agent, followed by polyamine encapsulation and PEG modification to obtain composite nanoparticles for synergistic cancer magnetic hyperthermia and SDT. The Fe₃O₄ component exhibited catalase activity, facilitating oxygen production within the TME rich in hydrogen peroxide and alleviating cancer hypoxia to improve the SDT effect.²⁶¹ Yang et al. synthesized DSPE-PEG2000 modified hyperfine photoetched bismuth vanadate (BiVO₄) nanorods (PEBVO@PEG NRs) capable of in situ O₂ and ROS supply for hypoxic cancer therapy. Experimental results showed that PEBVO@PEG effectively overcame the main obstacles faced by traditional sonosensitizing agents during SDT and exhibited significant SDT efficacy, presenting a new strategy to enhance the sonosensitizing agent performance and enable hypoxic cancer treatment [Fig. 10(c)].²⁶² 

Combining chemotherapy and immunotherapy offers several advantages, including synergistic therapeutic mechanisms, reduced drug dosage, and enhanced therapeutic effect, making it a promising approach for cancer treatment. However, challenges arise from the distinct pharmacokinetic properties and in vivo distribution of the two agents, inconsistent targets, uncontrolled drug ratios at cancer sites, and severe systemic side effects. To overcome these limitations and achieve better synergistic anti-cancer effects, nanodrug delivery systems have emerged with their improved pharmacokinetic properties, cancer-targeting delivery capabilities, and responsive targeted drug delivery in the TME. These systems facilitate the delivery of chemotherapeutic immunotherapeutic agents.²⁶³ Hu et al. designed polyacrylate-coated ultra-small superparamagnetic iron oxide (PAA@IONs) and dipeptide-modified liposomes co-loaded with DOX and β-lapafenone (Lap) (NRDL-Lip) to simultaneously target triple-negative breast cancer (TNBC) cells and tumor-associated macrophages (TAM). The results showed that PAA@IONs promoted the polarization of macrophages into anti-cancer M1-TAM within the cancer and immunosuppressive TME. This action reversed the TME and consistently elicited a robust immune response, inhibiting organ-specific metastasis during the progression of TNBC. NRDL-Lip enabled the accumulation and release of DOX and Lap at the cancer site, leading to the reversal of drug resistance, induction of robust ICD, and reshaping of the local immunosuppressive TME. This innovative strategy provides a new chemoimmunotherapy option for TNBC treatment [Fig. 11(a)].²⁶⁴ 

In recent years, photothermal therapy has been shown to induce ICD in cancer cells, leading to enhanced innate and adaptive immune responses in preclinical models. In addition, photothermal therapy improves cancer perfusion and reduces interstitial pressure, possibly by transiently increasing vascular permeability. This promotes the accumulation of therapeutic costimulatory molecules, immune cells, and the release of proinflammatory cytokines and chemokines that can overcome immunosuppressive TME.²⁶⁵ Various clinical trials combining photothermal therapy with immunotherapy have shown promise in advancing cancer treatment. BP nanosheets, as metal-free photothermal agents, can efficiently convert NIR light into heat energy. Xie et al. explored the role of PTT using two-dimensional BP in improving a CD47-mediated ICB therapy. PTT mediated by BP nanoparticles not only directly destroys cancer cells but also improves the inherent low immunogenicity of cancer cells. When combined with anti-CD47 therapy, it induces differentiation of TAMs into the M1 phenotype, improves cancer-specific antigen presentation by macrophages, and stimulates the generation of cancer antigen-specific T cells. These T cells can migrate to distant cancers, effectively suppressing metastatic cancers. Moreover, the impact of MSNs on photothermal immunotherapy has also been investigated.²⁶⁶ Yue et al. developed mesoporous hexagonal core–shell zinc porphyrin silica nanoparticles (MPSNs) containing imiquimod (R837, TLR7 agonist) for combination therapy involving PDT, PTT, and specific immunity in breast cancer. The nanoparticle design utilizes zinc tetraphenylporphyrin as the core and MSNs as the shell. The particles efficiently produce ROS and convert photons into heat energy using a single light source, achieving PDT and PTT effects. The mesoporous structure of the silica shell enables efficient loading of R837, which collaborates with cancer-related antigens, effectively promoting the maturation of dendritic cells and stimulating a strong immune response. MPSNs can be used as excellent photosensitizers and efficient drug carriers. The therapeutic strategy based on MPSNs@R837 not only eliminates primary cancers through PDT and PTT but also effectively inhibits cancer metastasis by inducing a strong bidirectional mechanistic immune response [Fig. 11(b)].²⁶⁷ 

The combination of phototherapy and immunotherapy has emerged as a promising approach for the treatment of primary and metastatic cancers. Increasing evidence supports the ability of PDT to induce ICD. By utilizing NIR light, PDT activates photosensitizers to produce ROS to kill cancer cells. Furthermore, ROS-induced endoplasmic reticulum stress leads to the release of calcium reticulum proteins, triggering the release of TAAs and activating immune responses. The integration of immunotherapy with phototherapy has given rise to a novel anti-cancer strategy called photoimmunotherapy.²⁶⁷ Wu et al. designed an immunostimulatory called CC@SiO₂-PLG for synergistic cancer therapy with photoactivated immunotherapy capabilities. This platform consists of catalase and a photosensitizer encapsulated within a silicon dioxide capsule, with an immune stimulator linked to it via an oxygen-cleavable linker. Upon accumulation of CC@SiO₂-PLG in cancer tissue, laser irradiation induces the production of O₂, alleviating cancer hypoxia and promoting singlet oxygen generation. This process not only destroys the cancer but also releases TAAs. Simultaneously, singlet oxygen generated by light damages the linker, triggering the remote release of an indoleamine 2,3-dioxygenase (IDO) inhibitor conjugated to CC@SiO₂-PLG. This reverses the immunosuppressive TME. The released TAAs, combined with the inhibition of the IDO-mediated tryptophan/canine uric acid metabolic pathway, induce a potent anti-cancer immune response induced by CC@SiO₂-PLG-mediated phototherapy. In a mouse xenograft cancer model, this approach significantly inhibits the growth of primary and distant cancers, as well as lung metastases, surpassing the effects achievable by phototherapy alone [Fig. 11(c)].²⁶⁸ 

Combining photothermal chemotherapy with cancer therapy offers a synergistic approach that capitalizes on the strengths of each treatment modality. Extensive research has been conducted in this field. He et al. constructed a composite nanoparticle consisting of GNRs and DNA tetrahedra, enabling the targeted delivery of a large quantity of chemotherapy drugs to cancer cells. The photothermal effect of GNRs complements the chemotherapy aspect, resulting in combined photothermal chemotherapy. The surface of GNRs was modified with thiolated PEI via an Au–S bond, enhancing the cellular uptake and escape from lysosomes. By utilizing base pairing principles, a DNA tetrahedron modified by the nucleic acid aptamer AS1411 was assembled and incubated with the anti-cancer drug DOX to obtain a high drug-loading DNA tetrahedron. Composite drug-loaded nanoparticles were obtained by electrostatically encapsulating the DNA tetrahedron within the PEI layer on the GNR surface. Each AS1411-functionalized DNA tetrahedron can carry DOX drug molecules, exhibiting strong targeting effects on HeLa and 4T1 cells. It enables accurate drug delivery to cancer sites, enhancing drug delivery efficiency. The PEI layer on the GNRs compresses the DNA drug carrier, facilitating intracellular and lysosomal escape through the proton sponge effect. Upon NIR laser irradiation, the composite drug-loaded nanoparticles induce apoptosis and necrosis of cancer cells, demonstrating the superior anti-cancer efficacy resulting from the synergy of PTT and chemotherapy. Nanocomposite delivery platforms based on GNRs and DNA tetrahedrons hold significant potential for cancer therapy [Fig. 12(a)].²⁶⁹ Deng et al. designed a drug delivery nanosystem based on BP QDs for targeted chemo-PTT. To achieve effective targeting, they electrostatically attached FA to the surface of BPQDs. Subsequently, DOX was loaded onto the FA-functionalized BP QDs for chemotherapy. Under NIR laser irradiation, the drug delivery system exhibited potent cancer cell-killing ability [Fig. 12(b)].²⁷⁰ Niu et al. designed a multi-functional complex called AuNR@PDA by modifying polydopamine (PDA) on the surface of AuNRs and connecting microRNA-192, which inhibits autophagy gene expression. This design greatly improved the biocompatibility, photothermal effect, and thermal stability of AuNR@PDA, enabling combined gene therapy and PTT.²⁷¹ 

PDT is widely used as an efficient, non-invasive, and low-toxicity treatment method. However, the inability of photosensitizers to target cancer aggregation and hypoxia within the cancer region limits the widespread application of PDT in cancer therapy. Combining PDT with other cancer treatments has shown promise in achieving improved treatment effects.^272,273 The concentration of chemotherapy drugs and oxygen at the cancer site is crucial for the efficacy of combined chemotherapy and PDT for anti-cancer therapy. Wang et al. addressed this challenge by utilizing the unique cavity and mesoporous pore structure of hollow MSNs to simultaneously load O₂-saturated perfluoropentane (PFP), ICG, and the chemotherapy drug DOX. These drug-loaded nanoparticles were then coated with poly(dopamine), which possesses photothermal conversion characteristics, resulting in the creation of a delivery system with high drug-loading, oxygen supplementation, and drug release promotion characteristics called HFODI@P. Under 808 nm light irradiation, the temperature of the drug carrier system and the TME increases, leading to a liquid–gas phase transition of PFP. The dual effects of PFP gasification and temperature increase trigger the explosive release of DOX and O₂, thereby enhancing the combined anti-cancer effect of chemotherapy and PDT. In addition, combining PTT, PDT, and immunotherapy has demonstrated promising results [Fig. 13(a)].²⁷⁴ Li et al. developed Bi₂Se₃@AIPH nanoparticles with high photothermal conversion capability by encapsulating the free radical initiator azodiisobutylimidazoline hydrochloride (AIPH) and the phase transition material lauric acid within hollow bismuth selenide nanoparticles (Bi₂Se₃NPs). These nanoparticles can achieve PTT, oxygen-independent PDT, and immunotherapy in a cascade manner under single 808 nm laser irradiation. Upon injection of the nanoparticles into the tail vein of cancer-bearing mice, 808 nm NIR light irradiates the cancer area, triggering excessive heat. This heat directly kills cancer cells while accelerating the release of AIPH and stimulating the production of toxic free radicals, further contributing to cancer cell death. The results demonstrated significant inhibition of cancer growth, with some cancers even completely disappearing and achieving a cancer growth inhibition rate of up to 99.6% [Fig. 13(b)].²⁷⁵ 

Despite the great potential of functionalized inorganic nanoparticles in cancer diagnosis and treatment, there are significant challenges for inorganic nanoparticles to effectively applying these materials in clinical settings. The construction of functional inorganic nanoparticles requires the use of different modification materials, and inorganic nanoparticles modified with similar or identical materials have significantly different effects. Even minor differences in preparation conditions can result in significant differences in properties, morphology, and effects. The choice of modification methods, connection methods, and their combination, recombination, and proportion adjustment have a crucial impact on the efficacy of cancer treatment. There is ample room for improvement and adjustment in the synthesis and modification of inorganic nanomaterials. One major challenge is the biosafety of inorganic nanomaterials, which remains a significant concern for their successful clinical application. Although efforts have been made to minimize the toxicity of functionalized inorganic nanoparticles, there is still a lack of clear evidence regarding their effective metabolism in the body. Inorganic materials cannot be rapidly metabolized after systemic administration, leading to their accumulation in the reticuloendothelial system, and potentially causing long-term toxicity, inflammatory reactions, fibrosis, or even cancer. Moreover, determining the safe dose level of inorganic nano-based drugs is challenging due to the lack of clear evaluation criteria. Furthermore, inorganic nanomaterials encounter several issues during application. For example, materials such as BP are susceptible to degradation in the physiological environment. To prevent degradation, BP needs to be coated with a polymer or a cell membrane, which adds complexity to the overall drug delivery system and hinders the clinical application of BP nanomaterials. In addition, inorganic nanoparticles often exhibit low photothermal conversion efficiency and poor water dispersibility and can induce immune responses. Their high cost, primarily attributed to the presence of heavy metals and rare metals, further impedes their clinical application.

Cancer poses a significant global health threat, and effective diagnosis and treatment measures are still lacking. However, nanotechnology has provided promising advancements in utilizing functional inorganic nanomaterials for cancer diagnosis and treatment. In this Review, we explore the unique characteristics and modification strategies of various types of inorganic nanoparticles. The functional modification of inorganic nanoparticles can mainly include the adjustment of size, shape, and potential; surface modification of organic nanomaterials; surface modification of inorganic nanomaterials; and surface modification of functionalized components. Targeted modification can enable inorganic nanoparticles to recognize specific structures or targets in cancer tissue and cells and complete specific delivery of the encapsulated drugs. By designing stimulation-responsive inorganic nanoparticles based on cancer characteristics, controlled drug release at the cancer site can be achieved, while also improving immune system recognition, thus synergistically enhancing treatment efficacy. In addition, the application of functionalized inorganic nanoparticles in cancer biomarker detection and imaging is described. Functionalized inorganic nanoparticles have excellent in vivo imaging capabilities, benefiting from their strong NIR optical absorption and x-ray attenuation capabilities for improved detection. Finally, the advantages and toxicity of functional inorganic nanoparticles are discussed in the context of various cancer treatment methods, such as radiotherapy, chemotherapy, gene therapy, immunotherapy, PTT, PDT, SDT, and combination therapy. Inorganic nanoparticles have been widely used in cancer diagnosis and treatment and have made significant progress, leveraging their photothermal and photodynamic effects, drug-loading capacity, biocompatibility, and biodegradability.

However, successfully translating functionalized inorganic nanoparticles to clinical applications requires further effort. It mainly includes the following aspects: (1) Improving the targeting and controllability of inorganic nanomaterials to cancers, enhancing the concentration of drugs at cancer sites, and minimizing toxic effects on normal tissues. (2) Reducing the toxicity of functional inorganic nanomaterials to meet clinical application standards. (3) Exploring better inorganic nanomaterials, improving biocompatibility and biosafety, and designing specific carriers for different drug solubilities to enhance cellular uptake and utilization. (4) Understanding the functional conditions, mechanisms, and degradation processes of inorganic nanomaterials in organisms. With ongoing advancements in nanotechnology, functional inorganic nanomaterials will be increasingly used in clinical medicine.

This work was supported by the National Natural Science Foundation of China (Grant No. 82173341) and the Natural Science Foundation of Hunan Province (Grant No. 2021JJ40845).

The authors have no conflicts to disclose.

M.L., Q.T., L.L., and S.L. wrote the main manuscript text, and H.W., G.Z., and D.Y. prepared the figures. All authors reviewed the manuscript.

Mengmeng Li: Conceptualization (equal); Writing – original draft (lead). Qinglai Tang: Conceptualization (equal); Validation (lead). Hua Wan: Conceptualization (equal); Supervision (equal). Gangcai Zhu: Funding acquisition (lead); Supervision (equal); Validation (equal). Danhui Yin: Conceptualization (supporting); Visualization (equal). Lanjie Lei: Conceptualization (equal); Validation (equal); Visualization (equal). Shisheng Li: Conceptualization (equal); Funding acquisition (equal); Supervision (equal); Validation (equal).

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