Titanium Dioxide-Based Nanoparticles to Enhance Radiation Therapy for Cancer: A Literature Review

Abstract

Titanium dioxide nanoparticles (TiO₂ NPs) have been investigated as one of the potential dose enhancement agents for radiation therapy. The role of TiO₂ NPs as a photodynamic sensitiser has been well documented, but its sensitisation with X-rays is not highlighted. Unlike other metal NPs, such as gold NPs, the main challenge for TiO₂ NPs as radiosensitisers is their low atomic number, resulting in a small cross-section for X-rays. This review summarises the results of current research in this area to explore the dose enhancement inflicted by TiO₂ NPs, which could potentially be of great value in improving radiation therapy efficiency.

1. Introduction

Radiation therapy using X-rays is currently one of the most common ways to treat cancer, which is the leading cause of death worldwide [1,2], and approximately half of all cancer patients receive radiation therapy during their treatment course [2,3]. X-rays can excite and ionise water molecules in the body, which subsequently induces free radicals and reactive oxygen species (ROS) [4]. Some radicals, such as the hydroxyl radical and the superoxide anion radical, are highly reactive and can damage intracellular DNA and subcellular structures such as membranes, mitochondria, etc., leading to cell death [5,6,7]. Thanks to the development of technologies, the current clinical practice of radiation therapy, including stereotactic body radiation therapy, intensity-modulated radiation therapy, and image-guided radiation therapy, is able to minimise the irradiated volumes with high geometrical accuracy [8], allowing the prescribed dose of radiation to be escalated to cancer cells while sparing the surrounding normal cells. Despite these advances in medical physics, overcoming radioresistance or advanced-stage tumours remains a challenging task for radiation therapy. For example, pancreatic cancer is a highly lethal tumour that is often diagnosed at an advanced stage and is also resistant to chemotherapy and radiation therapy [9]. Therefore, novel biological and chemical strategies in combination with radiation therapy are warranted to improve the treatment outcomes of this type of tumour.

Metal-based nanoparticles (NPs) have been extensively studied as potential radiosensitisers in radiation therapy over the last two decades. They can generate secondary electrons and, consequently, ROS by interacting with X-rays in tumour tissues, which can enhance the cellular damage caused by X-rays [6]. Their radiosensitising properties may increase the effect of X-rays on cancer cells while minimising the dose to healthy tissues. In addition, the absorption of X-rays by NPs can also be exploited as a potential agent for enhancing image contrast in computed tomography (CT) [10,11]. While a number of metallic materials have been used for radiosensitising NPs, gold NPs (Au NPs) are one of the most commonly used nanomaterials, and several literature and review articles have been published [12,13,14,15].

Titanium dioxide NPs (TiO₂ NPs) have been widely investigated as photodynamic therapy (PDT) applications for cancer treatment due to their photosensitising effects. In response to ultraviolet (UV) light, which has a longer wavelength than X-rays, the electrons in the TiO₂ crystal are excited, generating a plethora of ROS via the photocatalysis process of water molecules in tissues [16,17,18]. In vitro and in vivo studies have demonstrated the potential effects of UV-activated TiO₂ NPs against cancer cells in PDT [19,20]. However, unlike conventional radiation therapy using high-energy X-rays, UV light cannot penetrate the human body to reach deep-seated tumours such as pancreatic cancer, limiting the applications of TiO₂ NPs as PDT to superficial.

In contrast to the investigations of TiO₂ NPs as PDT applications with UV light, fewer studies have been reported using TiO₂ NPs for radiation therapy, where X-rays are commonly used. As X-rays have a rather higher photon energy than that of UV light, the photocatalytic activities of TiO₂ by X-rays may not be efficient but may still be induced by the decay process of excited titanium atoms [21]. In terms of the physical reactions, the dominant effects of TiO₂ NPs with high-energy photons could be a photoelectric effect and Compton scattering. However, one of the challenges of using TiO₂ NPs as radiosensitisers is the lower atomic number (Z = 22) and mass attenuation coefficient (µ/ρ = 0.2721 cm²/g) compared to other metal NPs such as gold (Z = 79 and µ/ρ = 5.158 cm²/g at 100 keV photons) [22], resulting in less interaction of NPs with X-rays, especially in the kV energy range. To date, the effect of TiO₂ NPs on radiation dose enhancement has not been well documented or concluded. It is also possible that dose enhancement using TiO₂ NPs could be achieved by modifying them with other chemicals to improve the interaction with X-rays. This review aims to summarise the literature on the use of TiO₂-based NPs with X-rays to identify their potential as dose enhancement agents in radiation therapy.

2. Materials and Methods

Literature searches were conducted in PubMed and Elsevier ScienceDirect using combinations of search terms including ‘titanium dioxide nanoparticles’, ‘X-rays’, ‘radiation’, ‘radiation therapy’, and ‘cancer’. The search was restricted to English-language texts published up to June 2022. All abstracts were screened for relevance and included articles in which in vitro and/or in vivo experiments were performed using TiO₂-based NPs and high-energy X-rays (kV and MV energy range) with any cell lines, even if the main objective of the study did not focus on the dose enhancement effects of NPs. A particle size of less than 100 nm was included as an NP. Phantom studies of TiO₂-based NPs using CT were also considered relevant, as kV X-rays are used, whereas studies using X-rays for structural analysis of TiO₂ NPs, for example, X-ray absorption fine structure analysis, were excluded as they were not intended for radiation therapy. Simulation studies and review articles were also excluded. Screening of titles and abstracts against the inclusion criteria yielded 54 articles. The full text of these articles was read, and finally, a total of 25 articles, including seven articles from our associate group, were selected as directly relevant citations for the purpose of our review.

3. Radiation Dose Enhancement Effect of TiO₂ NPs

As there are different types of TiO₂ NPs with different sizes, shapes, and surface modifications or dopants used in various studies, we categorised them into two types to show their dose enhancement effects: unconjugated and conjugated TiO₂ NPs. The conjugated TiO₂ NPs were referred to as TiO₂-based NPs that were modified with other materials to increase their dose enhancement effect in response to X-rays, such as doping gold into TiO₂ NPs [23], while the unconjugated TiO₂ NPs did not contain such materials in the NPs. If no specification was given in an article, it was assumed to be unconjugated TiO₂ NPs.

The dose enhancement effects of TiO₂ NPs in vitro using cancer cell lines were investigated in 13 studies, each for the unconjugated and conjugated TiO₂ NPs. However, it was found to be challenging to directly compare their results following a unified method because the experimental designs, methods, and conditions were quite different between the studies, including nanoparticle sizes, concentrations, surface modifications, X-ray energies, irradiated doses, cell lines, and cellular uptake of NPs. Therefore, we simply calculated a sensitiser enhancement ratio as the ratio of cell viability with NPs to that without NPs under X-ray irradiation at the maximum doses used in each experiment to estimate the effects of TiO₂ NPs across the literature and document it in this review. As the specific values of the cell viability were not reported in most of the articles, they were read off from each figure using WebPlotDigitizer (https://apps.automeris.io/wpd/, accessed on 1 March 2023). The sensitiser enhancement ratios for unconjugated and conjugated TiO₂ NPs for each literature are summarised in

Table 1. Summary of in vitro studies using unconjugated TiO₂ NPs.

Study [Ref]	Particle Size (nm)	Cell Line	Assay	Radiation Energy	NPs Concentration	Surface Modification/ Other	Enhancement Ratio
Akasaka et al. [24]	50	MIAPaCa-2	Clonogenic	150 kVp	100 µg/mL	N/A	2.63 @ 8 Gy
Bakhshizadeh et al. [25]	62.8	DFW, HT1080	MTT	6 MV	1.33 µg/mL	Methacrylic acid (MAA), Nanopolymer	1.10 @ 1 Gy (DFW) 0.85 @ 1 Gy (HT1080)
Cheng et al. [23]	50.6	SUM159	CellTiter-Blue	320 kVp	5 µg/mL	Polyethylene glycol (PEG)	1.14 @ 10 Gy
Gerken et al. [26]	5.2	HT1080	CellTiter-Glo	150 kVp, 6 MV	80, 160, 320 µg/mL	Flame spray pyrolysis synthesis	2.65 @ 8 Gy (320 µg/mL, 150 kVp) 2.30 @ 8 Gy (320 µg/mL, 6 MV)
Gerken et al. [27]	5.3	HT1080, HeLa	CellTiter-Glo	150 kVp	1 mg/mL	Flame spray pyrolysis synthesis	1.77 @ 8 Gy (HT1080) 0.61 @ 8 Gy (HeLa)
Mirjolet et al. [28]	10	SNB-19, U87MG	Clonogenic	Medical LINAC (Possibly 6 MV)	1 µg/mL	Nanotube-shaped	2.25 @ 5 Gy (SNB-19) 1.35 @ 10 Gy (U87MG)
Morita et al. [29]	50	BxPC3	Clonogenic	80 kVp	1 mg/mL	Polyacrylic acid (PAA)	1.70 @ 5 Gy
Nakayama et al. [30]	12.7	A549, DU145	Clonogenic	6 MV	200 µg/mL	Aminopropyl trimethoxysilane (APTS), Polyethylene glycol (PEG)	1.18 @ 6 Gy (A549, APTS) 1.24 @ 6 Gy (A549, PEG) 1.20 @ 6 Gy (DU145, APTS) 1.16 @ 6 Gy (DU145, PEG)
Ouyang et al. [31]	5	A549	Clonogenic	6 MV	0.5 µg/g	N/A	1.34 @ 2 Gy
Pan et al. [32]	24	B16-F10, 4T1-Luc	Clonogenic	Medical LINAC (Possibly 6 MV)	100 µg/mL	Cancer cell membrane	1.49 @ 4 Gy (B16-F10) 1.32 @ 4 Gy (4T1-Luc)
Rezaei-Tavirani et al. [33]	N/A	MCF-7, MKN-45	MTT	60Co (1.173, 1.332 MeV γ)	30 µg/mL	Anatase, Rutile TiO2	2.73 @ 2 Gy (MCF-7, Anatase) 0.87 @ 2 Gy (MCF-7, Rutile) 1.60 @ 2 Gy (MKN-45, Anatase) 0.58 @ 2 Gy (MKN-45, Rutile)
Su et al. [34]	11.78	SW1990	Clonogenic	125I (internal radiation source, 35.5 keV γ)	144 µg/mL	Oleic acid	2.07 @ 7.5 µCi
Youkhana et al. [35]	30	HaCaT, DU145	Clonogenic	80 kVp, 6 MV	4 mM	Aminopropyl trimethoxysilane (APTS)	8.00 @ 8 Gy (HaCaT, 80 kVp) 2.22 @ 8 Gy (HaCaT, 6 MV) 13.50 @ 8 Gy (DU145, 80 kVp) 1.90 @ 8 Gy (DU145, 6 MV)

N/A, not available; LINAC, linear accelerator.

and

Table 2. Summary of in vitro studies using conjugated TiO₂ NPs.

Study [Ref]	Conjugated Material	Surface Modification	Particle Size (nm)	Cell Line	Assay	Radiation Energy	NPs Concentration	Enhancement Ratio
Alban et al. [36]	Sodium (Na), Zinc (Zn)	Quercetin, Nanotube-shaped	9–12	T24	Clonogenic	60Co (1.173, 1.332 MeV γ)	25 µg/mL	1.75 @ 5 Gy (Na) 2.86 @ 5 Gy (Zn)
Bakhshizadeh et al. [25]	Mitoxantrone (MX)	Methacrylic acid (MAA), Nanopolymer	62.8	DFW, HT1080	MTT	6 MV	1.33 µg/mL	6.11 @ 1 Gy (DFW) 1.23 @ 1 Gy (HT1080)
Cheng et al. [23]	Gold (Au)	Polyethylene glycol (PEG)	70.1	SUM159	CellTiter-Blue	320 kVp	5 µg/mL	2.20 @ 10 Gy
Gao et al. [37]	Tungsten (W)	Polyethylene glycol (PEG)	9.1	4T1	Clonogenic	N/A	100 µg/mL	2.47 @ 8 Gy
Hassan et al. [38]	Hydrogen peroxide (H2O2)	Polyacrylic acid (PAA)	50	MIAPaCa-2	Clonogenic	150 kVp	400 µg/mL	4.58 @ 5 Gy
Morita et al. [29]	Hydrogen peroxide (H2O2)	Polyacrylic acid (PAA)	50	BxPC3	Clonogenic	80 kVp	1 mg/mL	35.5 @ 5 Gy
Nakayama et al. [39]	Hydrogen peroxide (H2O2)	Polyacrylic acid (PAA)	70	MIAPaCa-2	Clonogenic	150 kVp	0.15% w/v	3.40 @ 5 Gy
Nakayama et al. [30]	Samarium (Sm)	Aminopropyl trimethoxysilane (APTS), Polyethylene glycol (PEG)	12.7	A549, DU145	Clonogenic	6 MV	200 µg/mL	1.48 @ 6 Gy (A549, APTS) 1.54 @ 6 Gy (A549, PEG) 1.42 @ 6 Gy (DU145, APTS) 1.46 @ 6 Gy (DU145, PEG)
Pan et al. [32]	Manganese dioxide (MnO2)	Glucose oxidase (GOx), Cancer cell membrane	46	B16-F10, 4T1-Luc	Clonogenic	LINAC (Possibly 6 MV)	100 µg/mL	16.23 @ 4 Gy (B16-F10) 18.85 @ 4 Gy (4T1-Luc)
Pan et al. [40]	SN-38, TAT peptide	RGD peptide, Mesoporous-shaped	45	4T1-Luc	Clonogenic	6 MV	100 µg/mL	5.09 @ 4 Gy
Tekin et al. [41]	Platinum (Pt), Zirconium (Zr)	SPHINX	46.3	PC3, LNCaP, RWPE-1	WST	6 MV + Light	10 µg/mL	2.09 @ 5 Gy (PC3) 2.29 @ 5 Gy (LNCaP) 1.10 @ 5 Gy (RWPE-1)
Townley et al. [42]	Gadolinium (Gd), Erbium (Er), Europium (Eu)	Silica	60	MCF7, RH30	Trypan blue	250 kVp	225 nM	1.60 @ 3 Gy (MCF7, 10%Gd1%Er1%Eu) 2.94 @ 3 Gy (RH30, 10%Gd1%Er1%Eu)
Yang et al. [43]	Cerium (Ce)	N/A	15	A549	WST	80 kVp	10 mg	1.18 @ 0.13 Gy *

SN-38, 7-Ethyl-10-hydroxy-camptothecin; TAT, transactivator of transcription; RGD, arginine-glycine-aspartic; N/A, not available; LINAC, linear accelerator. * No control data were available for X-rays alone. The enhancement ratio was calculated using data from undoped TiO₂ NPs + X-rays as a control.

, respectively.

3.1. Unconjugated TiO₂ NPs In Vitro Studies

The unconjugated TiO₂ NPs had slightly enhanced cell killing effects with X-ray irradiation compared to X-ray irradiation alone in most studies, as shown in Table 1, and the enhancements were reported as significant in six studies. The overall mean sensitiser enhancement ratio for unconjugated TiO₂ NPs, calculated as the ratio of cell viability with NPs to that without NPs under X-ray irradiation in each study, was 2.2 with a large standard deviation of 2.7, indicating that the significant dose enhancement effect of unconjugated TiO₂ NPs in response to X-rays is controversial and dependent on the study. The clear trends or dependencies of particle size, concentration, X-ray energy, and irradiated dose on the sensitiser enhancement ratio could not be found among the studies. The maximum radiosensitising effect of unconjugated TiO₂ NPs was observed in a study carried out by Youkhana et al. [35]. Their synthesised TiO₂ NPs, the surfaces of which were modified with aminopropyl trimethoxysilane, showed significant radiosensitisation to human keratinocyte (HaCaT) and prostate (DU145) cancer cell lines using 80 kV X-rays at a dose of 0–8 Gy. Moreover, they performed the clonogenic assay using both kV and MV X-rays and reported that the dose enhancement effect of TiO₂ NPs with 80 kV X-rays was greater than that with 6 MV X-rays due to the different dominant process of the interaction of X-ray photons with titanium atoms in the NPs. Contrary to this result, comparable dose enhancement effects for both kV and MV X-rays using human fibrosarcoma HT1080 cells were found in other work reported by Gerken et al. [26]. They suggested that, rather than the physical reactions, the catalytic reactions of TiO₂ NPs with X-rays played an important role in the generation of ROS and cellular oxidative stress.

There have been several studies showing that TiO₂ NPs have no radiosensitising effect on some types of cancer cell lines. Another study reported by Gerken et al. [27] evaluated cell viability using HT1080 and HeLa cell lines with 150 kVp X-rays and showed that there was no decrease in the surviving fraction in HeLa cells treated with TiO₂ NPs compared to the control groups, while the decrease was found in HT1080 cells. Their result suggests that the dose enhancement effects of TiO₂ NPs may be cell line-dependent due to the different radiosensitivity for different cell types. In another study using polymer-structural TiO₂ NPs, no radiation sensitivity of TiO₂ NPs was observed in HT 1080 and DFW melanoma cell lines under exposure to 6 MV X-rays [25]. This might be due to the high energy of X-rays and also due to the polymer structure, which could absorb the generated secondary electrons and ROS. In general, higher dose enhancements are recorded for low-energy (kV) X-rays, whereas other factors, including particle size, shape, concentration, and synthesis method, may affect the dose enhancement.

3.2. Conjugated TiO₂ NPs In Vitro Studies

The dose enhancement effect appears to be more pronounced in the TiO₂ NPs modified with other materials (referred to in this review as “conjugated TiO₂ NPs”) than in the unconjugated TiO₂ NPs. Eleven studies showed significant X-ray dose enhancements on cell survival using their modified TiO₂ NPs. Two studies by Pan et al. did not present the statistical analysis data in their articles, but the dose enhancement effects are obvious, as the surviving fraction for cells treated with their conjugated TiO₂ NPs and X-rays was less than 20%, whereas it was more than 80% for cells treated with X-rays alone in both studies [32,40]. A variety of materials have been used to improve the efficacy of TiO₂ NPs as radiosensitisers, as summarised in Table 2. Based on the different stages of radiation action, i.e., the physical, chemical, and biological interactions with NPs, the strategies of conjugating materials to improve the dose enhancement effects of TiO₂ NPs were mainly classified into three types as follows: (1) to increase physical interactions of NPs with X-rays using different metallic materials; (2) to generate additional ROS by modification with other chemical compounds; and (3) to control individual cellular responses using ligand modalities such as a cell-targeting peptide. These approaches are illustrated in Figure 1 and discussed in detail below.

3.2.1. Dose Enhancement by Physical Approaches

Doping other metallic materials with TiO₂ NPs could be one of the most common strategies to physically improve their radiosensitivity. These dopants have a higher atomic number than that of titanium, resulting in more interactions of the NPs with X-ray photons and consequently more secondary electrons generated via photoelectric and Compton effects. Townley et al. incorporated gadolinium and other rare earth elements into the silica-coated TiO₂ NPs to optimise the X-ray absorption of the NPs and showed that the doped TiO₂ NPs had a significant reduction in cell proliferation at 3 Gy of irradiation in the MCF7 and RH30 rhabdomyosarcoma cell lines [42]. They also reported that there was statistically insignificant but reduced cell growth in an embryonal rhabdomyosarcoma with the doped TiO₂ NPs and X-ray irradiation. Similarly, samarium, which is one of the rare earth elements, has been investigated as a doped material in TiO₂ NPs and showed significant radiation dose enhancement compared to undoped TiO₂ NPs [30]. Cerium appears to be another rare earth element used in doping. Cerium-doped TiO₂ NPs have shown lower cell viability in A549 lung cancer cells with 80 kV X-ray than that of undoped TiO₂ NPs [43]. Tungsten-doped TiO₂ NPs have also revealed significant radiosensitisation using 4T1 breast cancer cells [37]. Another study explored the use of Au NPs in conjunction with TiO₂ NPs and found significant cell death in SUM159 breast cancer cells with Au-TiO₂ NPs more than with either Au NPs or TiO₂ NPs or a simple mixture of them at 10 Gy of 320 kVp X-ray irradiation [23]. The result suggested synergistic effects from the combination of Au and TiO₂ NPs.

3.2.2. Dose Enhancement by Chemical Approaches

Conjugation with chemical compounds to generate additional ROS could be another prospect to improve the dose enhancement effects of TiO₂ NPs under X-ray irradiation. For example, hydrogen peroxide (H₂O₂)-modified TiO₂ NPs have been studied by a group of Morita et al. [29,38,39]. Interestingly, they found that the TiO₂ NPs could confine H₂O₂ molecules and release them into local tissue, which contributed to the radiosensitisation of TiO₂ NPs on cells as the main ROS agent. Their results showed that H₂O₂-modified TiO₂ NPs caused higher dose enhancement effects than those of unmodified TiO₂ NPs in pancreatic cancer cells. In another study by Pan et al., glucose oxidase decorated TiO₂ NPs with manganese dioxide (MnO₂) shell and cell membrane coating had a strong cell inhibition effect in mouse melanoma (B16-F10) and breast cancer (4T1-Luc) cell lines treated with 4 Gy of X-rays [32]. After glucose oxidase converted cellular glucose into intratumoural H₂O₂, MnO₂ acted as a catalyst to produce O₂ by H₂O₂, and TiO₂ NPs could effectively generate ROS using O₂ under X-ray irradiation. Conjugation of a photosensitiser into TiO₂ NPs was also investigated to extend the application of the TiO₂ NPs PDT technique to radiation therapy [25]. When NPs are irradiated by X-rays, a radioluminescence process occurs, which can activate the photosensitiser and induce additional ROS, resulting in enhanced cell killing through the same process as PDT. A reported study using a nanopolymer containing TiO₂ cores imprinted with mitoxantrone as a photosensitiser showed a synergistic enhancement of 6 MV X-rays on the survival of DFW and HT1080 cells compared to TiO₂ NPs without mitoxantrone [25].

3.2.3. Dose Enhancement by Biological Approaches

To increase the therapeutic efficacy of NPs through cell-based biological approaches, multi-functional TiO₂ NPs have been explored in a few studies. Pan et al. developed mesoporous TiO₂ NPs using 7-ethyl-10-hydroxy-camptothecin (SN-38) for cell cycle regulation in the G2/M phase and the arginine-glycine-aspartic acid (RGD) peptide for nuclear targeting [40]. The results showed that the nuclear-targeting TiO₂ NPs and 6 MV X-rays could effectively inhibit the growth, proliferation, and migration of 4T1-Luc cells. Targeting prostate cancer cells, Tekin et al. conjugated SPHINX, an inhibitor of serine-/arginine-rich protein kinase 1, which was overexpressed in prostate cancer, into platinum (Pt)-combined TiO₂ NPs [41]. Their hybrid TiO₂ NPs also targeted the intrinsic anti-angiogenic ability of SPHINX and the photosensitising property of Pt. They demonstrated a dramatic decrease in the cell viability of PC3 and LNCaP prostate cancer cells with radiation and light exposure, although they did not report the radiosensitising effect of the NPs without light exposure.

3.3. Dose Enhancement Effect on Normal Cells

Only one in vitro study has demonstrated radiosensitisation of normal cells by TiO₂ NPs. Wang et al. investigated the cytotoxicity of TiO₂ NPs and γ-rays from a ¹³⁷Cs source (662 keV) on human normal bronchial epithelial cells (16HBE14o-) [44]. As far as we have analysed the data they presented in the article, slight but probably insignificant dose enhancement effects were observed with 100 µg/mL of TiO₂ NPs and 4 Gy radiation. The sensitiser enhancement ratio we calculated was 1.16. They also performed the same experiments with silicon and aluminium, and both of them coated TiO₂ NPs, showing that the differences between the coated and uncoated TiO₂ NPs were small. The calculated sensitiser enhancement ratio was 1.07 for silica-aluminium-coated TiO₂ NPs at 4 Gy irradiation. It should be noted that upon cancer cell-targeting NPs, very small amounts are anticipated to be present in the surrounding healthy tissues, i.e., normal cells.

3.4. In Vivo Studies

Nine publications are reported in the literature on the radiosensitising effects of conjugated TiO₂ NPs in vivo using tumour-bearing mice treated by an intravenous or intratumoural administration of TiO₂ NPs and single fractional kV or MV X-ray irradiation, except in two studies where multiple fractions were used. Most of the studies assessed tumour growth after treatment with NPs and X-rays, and three studies reported a survival rate. Their results are summarised in

Table 3. Summary of in vivo studies using TiO₂ NPs for radiation therapy.

Study [Ref]	Conjugation/ Modification	Particle Size (nm)	Animal	Cell Line	Radiation Energy	Radiation Dose	NPs (Ti) Conc.	NPs Injection	Observation Period	Results
Cheng et al. [23]	Au/PEG	70.1	Mice	SUM159	320 kVp	10 Gy	4 mg/kg	100 µL, intravenous	37, 81 days	The tumour size treated with Au-TiO2 NPs + X-rays was fourfold smaller than that with X-rays alone. The median survival significantly increased.
Gao et al. [37]	W/PEG	9.1	Mice	4T1	N/A	4 Gy	3 mg/mL	200 µL, intravenous	14 days	The tumour growth rate of the group treated with W-doped TiO2 NPs + X-rays was lower than that of the X-rays alone group.
Hassan et al. [38]	H2O2/PAA	50	Mice	MIAPaCa-2	150 kVp	5 Gy	1.5 mg/mL	100 µL, intratumoral	55 days	H2O2-modified TiO2 NPs + X-rays showed significantly higher tumour growth inhibition compared with that of X-rays alone.
Nakayama et al. [39]	H2O2/PAA	70	Mice	MIAPaCa-2	150 kVp	5 Gy	8.7% w/v	150 µL, intratumoral	43 days	The tumour volume treated with H2O2-modified TiO2 NPs + X-rays was 35.4% of that with X-rays alone.
Pan et al. [32]	MnO2/GOx, Cell membrane	46	Mice	B16-F10, 4T1-Luc	LINAC (Possibly 6 MV)	4 Gy	1.2 mg/mL	150 µL, intravenous	40 days	The survival rate of the TiO2@MnO2-GOx@C + X-rays group was significantly prolonged compared to other groups.
Pan et al. [40]	SN-38, TAT/ RGD peptide	45	Mice	4T1-Luc	6 MV	6 Gy	60 mg/kg	150 µL, intravenous	21 days	Mesoporous TiO2(SN-38)-TAT-RGD NPs + X-rays greatly suppressed tumour growth.
Su et al. [34]	Oleic acid, 125I	11.78	Mice	SW1990	125I (internal radiation source, 35.5 keV γ)	600 µCi	144 µg/mL	10 µL, intratumoral	20, 60 days	The relative tumour volume treated with 125I-TiO2 NPs was 54.21% of that with 125I. The survival rate was significantly improved with 125I-TiO2 NPs.
Townley et al. [45]	Gd, Eu, Er/ Silica	65	Mice	A549	200 kVp	2 Gy × 5, 2.5 Gy × 10 + 2 Gy × 3	0.05, 1, 5 mg/mL	50 µL, intratumoral (days 0, 13, 20)	22 days	The tumours treated with 5%Gd-1%Eu-1%Er TiO2 NPs + X-rays were about half the size of those treated with X-rays alone.
Yang et al. [43]	Ce	15	Mice	A549	80 kVp	N/A (Multi-fraction)	N/A	N/A	10 days	The tumour size treated with Ce-doped TiO2 NPs + X-rays decreased significantly to 9.65% of the initial size.

N/A, not available; LINAC, linear accelerator.

. Each study used their unique conjugated TiO₂ NPs to enhance the radiosensitising effect of the NPs, resulting in significant tumour growth inhibition and/or prolonged survival for mice treated with TiO₂ NPs and X-rays compared to those treated with X-rays alone. While no studies focused on the unconjugated TiO₂ NPs in vivo, a survival profile of B16-F10 tumour metastasis mice treated by TiO₂ NPs coated with a cancer cell membrane and 4 Gy of X-rays was presented in a figure from a study by Pan et al. [32], showing that there was no difference in survival compared to those treated with X-ray irradiation alone. In the absence of in vivo studies reporting on the potential efficacy of unconjugated TiO₂ NPs as radiosensitsers, there could be no significant dose enhancement effects of TiO₂ NPs without modification.

4. ROS Generation by TiO₂ NPs Exposed to X-rays

Several in vitro studies have evaluated X-ray-induced ROS generation in TiO₂ NPs, which is considered to be one of the main sources of cellular damage. In practice, fluorescent dye-based assays are commonly used to measure ROS with or without cells due to the short lifetime of ROS generated by X-ray irradiation [6]. The results of ROS measurements reported in the literature appear to vary depending on the experimental setup. Following similar trends to those of cell survival measurements, ROS generation is shown to be increased in some studies with the combination of TiO₂ NPs and X-rays compared to that of X-rays alone and not increased in other studies, as shown in

Table 4. Summary of in vitro ROS measurements for TiO₂ NPs in reaction with X-rays.

Study [Ref]	Nanoparticles	ROS Probes	Cells	Measurements	ROS Increase in NPs + X-rays
Bakhshizadeh et al. [25]	MX-imprinted TiO2 NPs (Nanopolymers)	Terephthalic acid	No cells	Spectrofluorometer	Increase
Cheng et al. [23]	Gold-composed TiO2 NPs	DPBF, DCFDA	SUM159	Flow cytometry	Insignificant increase in TiO2 NPs, Significant increase in Au-TiO2 NPs
Gerken et al. [26]	TiO2 NPs	H2DCFDA	No cells	Microplate reader	Increase
Hassan et al. [38]	H2O2-modified TiO2 NPs	APF, DHE, c-H2DCFDA	MIAPaCa-2	Microplate reader, Microscope	Significant increase
Mirjolet et al. [28]	Titanate nanotubes	DCFDA	SNB-19, U87MG	Flow cytometry	No increase
Nakayama et al. [30]	Samarium-doped TiO2 NPs	DCFDA	A549, DU145	Microplate reader	Insignificant increase in TiO2 NPs, Significant increase in Sm-TiO2 NPs
Nakayama et al. [39]	H2O2-modified TiO2 NPs	APF, c-H2DCFDA, HE	MIAPaCa-2	Microplate reader, Flow cytometry	No increase in TiO2 NPs Significant increase in H2O2-TiO2 NPs
Pan et al. [32]	MnO2/GOx-decorated TiO2 NPs	DCFDA	B16-F10, 4T1-Luc	Microscope	Increase
Pan et al. [40]	Mesoporous TiO2 NPs	DBZTC, DCFDA	No cells	Spectrofluorometer	Possibly increase (No data are available for X-rays alone)
Su et al. [34]	125I-labelled TiO2 NPs	HPF	SW1990	Microscope	Significant increase (NPs + γ-rays of 125I)
Townley et al. [42]	Rare earth element-doped TiO2 NPs	Coumarin hydroxylation, c-H2DCFDA	RH30	N/A	No increase in TiO2 NPs Significant increase in doped TiO2 NPs
Yang et al. [43]	Cerium-doped TiO2 NPs	DCFDA	A549	Microplate reader	Significant increase in Ce-TiO2 NPs
Youkhana et al. [35]	TiO2 NPs	DCFDA	No cells	Microplate reader	Significant increase

DPBF, 1,3-diphenylisobenzofurn; DCFDA, 2′,7′-dichlorofluorescein diacetate; APF, 3′-(p-aminophenyl) fluorescein; DHE, dihydroethidium; HE, hydroethidium; DBZTC, 2-chloro-1,3-dibenzothiazoline cyclohexene; HPF, 3′-(p-hydroxyphenyl) fluorescein; N/A, not available.

. Even in the studies that showed TiO₂ NPs could improve ROS generation under X-ray irradiation, the level of increase in ROS seemed to be small and insignificant [23,25,30,32]. The amount of ROS generated could be increased by conjugating materials/chemicals to TiO₂ NPs, which could release additional ROS in response to X-ray irradiation [30,32,38,42,43]. In contrast, though, there are some studies that showed no difference in ROS generation between the presence and absence of TiO₂ NPs under X-ray irradiation [28,39,42]. One of the studies by Mirjolet et al. found no additional intracellular ROS production by titanate nanotubes, which were tube-shaped TiO₂ NPs, in glioblastoma cell lines despite their radiosensitising effects, as indicated by a clonogenic assay [28]. They suggested the biological effects, including an amplification of cell cycle arrest, by internalising titanate nanotubes into cells. The possible effects of TiO₂ NPs on intracellular ROS generation, other than those induced by physical interactions with X-rays, have also been reported in several studies. Youkhana et al. reported the discrepancy between in vitro and phantom experiments in the radiosensitising effects of TiO₂ NPs under MV X-ray irradiation [35]. Their data revealed significant enhancements of ROS generation and cell killing, but no dose enhancement in the phantom using the PRESAGE dosimeter, suggesting that the manifestation of the biological effects of TiO₂ NPs, which are nonexistent in the phantom, could be specifically attributed to ROS generated in cells. Another study by Gerken et al. indicated that the radiocatalytic activity of TiO₂ might lead to the generation of more ROS rather than the physical interactions through their in vitro study, which showed X-ray energy independence of the dose enhancement of TiO₂ NPs [26]. Compared to Au NPs, no significant difference in ROS production was reported between TiO₂ and Au NPs in SUM159 breast cancer cell lines [23]. In another comparative study by Hassan et al., Au NPs induced more hydroxyl radicals under kV-X-ray irradiation compared to their H₂O₂-modified TiO₂ NPs [38]. However, their results of a clonogenic assay using pancreatic cancer cells showed the significant dose enhancement effects of H₂O₂-modified TiO₂ NPs, while there was no significant enhancement with the Au NPs and X-rays, suggesting an important role of H₂O₂ for the dose enhancement.

5. Contrast Enhancement in CT Images

Five studies based on the investigation of the feasibility of employing TiO₂ NPs as CT contrast agents have been documented in the literature. According to these studies, the values of Hounsfield units (HUs) are linearly increased with the concentration of TiO₂ NPs; however, no visible changes are observed in CT images even at a high concentration of 5.0 mg/mL TiO₂ NPs due to the low atomic number of titanium [24,30]. For example, Nakayama et al. demonstrated that the CT number of TiO₂ NPs with a concentration of 2.0 mg/mL was about 3 HU, which is almost the same as that of water [30]. In another report by the same group using H₂O₂-modified TiO₂ NPs, the image contrast at a concentration of 10 mg/mL could be improved depending on the CT window width and level [46]. Another study showed that TiO₂ NPs with a very high concentration of 231 mg/mL had a CT number of 281.5 HU [47]. This should be sufficient to enhance the contrast in CT images, but such a concentration is too high to apply in vitro and in vivo studies as the cellular cytotoxicity of NPs will exceed the threshold value for cell viability. Therefore, in order to render TiO₂ NPs as CT contrast agents, it will be necessary to conjugate them with other materials to enhance the CT image contrast. A study by Gao et al. evaluated the performance of tungsten-doped TiO₂ NPs as contrast agents in CT images and showed that the image contrast of doped TiO₂ NPs was equivalent to that of iopromide, which is a commonly used CT contrast agent in clinical practice [37]. They also showed the contrast-enhanced tumour in CT images using 4T1 tumour-bearing mice injected intratumourally with 30 µL of the tungsten-doped TiO₂ NPs. As the photoelectric effect occurs with increasing atomic number in kV energy photons, high Z materials conjugated to TiO₂ NPs can enhance the absorption of kV X-rays, which are used in CT, resulting in enhanced contrast in CT images.

6. Challenges and Future Perspectives

Despite the potential of TiO₂ NPs as radiosensitisers, the number of published studies for this purpose is limited compared to those using other metal-based NPs, such as Au NPs, with large variations in study design and results. One of the major challenges in exploring the efficacy of TiO₂ NPs for radiation therapy, as well as other metal-based NPs, is the large variables in terms of nanoparticle shape, size, surface modifications, concentrations, cell types, and radiation energy, with the different effects in the physical, chemical, and biological phases. These variables should be optimised to maximise the effect of radiation dose enhancement on cancer cells while minimising side effects on normal tissues. Moreover, the mechanisms of the radiation dose enhancement effect of TiO₂ NPs need to be clarified in order to distinguish the advantages of using TiO₂ NPs as radiosensitisers from other metal NPs. The photoelectric effect and Compton scattering can occur in response to X-rays with metal NPs, generating free radicals and ROS and increasing the cellular damage caused by X-rays to surrounding tissues. However, the low atomic number and mass attenuation coefficient of TiO₂ result in less interaction of the NPs with X-rays than those of high-Z elements such as gold and gadolinium. Rather than the physical reactions with X-rays, the biochemical properties of TiO₂ NPs may be a key mechanism and should be explored in further studies. The extension of the photocatalytic property of TiO₂ NPs to radiation therapy may also be another potential challenge for their use as radiosensitisers, as PDT has shown significant results in cancer treatment in previous literature [19]. Before TiO₂-based NPs can be translated into clinical practice, more extensive preclinical studies are required, including assessments of long-term cytotoxicity on normal organs and quality assurance of production as biomaterial applications.

7. Conclusions

To the best of the authors’ knowledge, this is the first review article on TiO₂ NPs focusing solely on their potential as dose enhancement agents for radiation therapy. Unlike their PDT effects, the number of studies investigating the radiosensitising ability of TiO₂ NPs for ionising radiation is still limited, with variable experimental conditions among the studies. Thus, it was difficult to perform a systematic analysis through the reviewed literature to clarify the dose enhancement effect of TiO₂ NPs; however, certain trends found here could highlight TiO₂ NPs as radiation dose enhancement applications for future work.

Within the reviewed literature, TiO₂ NPs were found to cause certain dose enhancements to cells with X-rays, but this effect could be a matter of controversy regarding their levels of dose enhancement. Any dependencies of particle size and concentration, modifications, irradiated X-ray energy and dose, and cell lines on the dose enhancement effects could also not be deduced because the experimental conditions, including cell types and assay methods, were different in each reported study. Similar to other metal NPs such as Au NPs, ROS induced by TiO₂ NPs upon their exposure to X-rays is considered one of the main sources contributing to their dose enhancement effects, which were measured in several studies besides cell survival. However, the effect of generated ROS on radiation enhancement is also controversial. It should be noted that the biochemical interactions of TiO₂ NPs have been suggested to play a more significant role in the generation of ROS in some studies. This may be a unique feature of TiO₂ NPs that differs from other high-Z material NPs [26,28,35]. One of the interesting results indicating these effects was reported by Gerken et al. [26]. They showed that the dose enhancement effects of TiO₂ NPs on HT1080 cells in an in vitro study using 6 MV X-rays were higher than those of Au NPs at the same concentrations as TiO₂ NPs. The photocatalytic activity of TiO₂ NPs can also be utilised to increase ROS generation under X-ray irradiation [25,44]. Conjugation of other materials into TiO₂ NPs may be necessary to improve the dose enhancement effects of TiO₂ NPs to be significant for radiation therapy. A variety of materials have been employed to enhance the physical, chemical, and biological stages of radiation actions in TiO₂ NPs. The incorporation of a high-Z material should be an effective way for TiO₂ NPs to improve the therapeutic enhancement as well as the diagnostic enhancement in CT images.

There is a tendency to publish only results that are statistically significant, meaning that some data where TiO₂ NPs do not show dose enhancement effects may be missing. In addition to this bias, we noticed that several research groups reported the abilities of their conjugated TiO₂ NPs as radiosensitisers without presenting data on their unconjugated TiO₂ NPs [36,37,40,41,42]. It has been speculated that the unconjugated TiO₂ NPs had no dose enhancement effects in their experiments, which may be the reason why they developed the conjugated TiO₂ NPs. In fact, no in vivo studies have been reported using unconjugated TiO₂ NPs with radiation therapy, as far as we have investigated.

Taken together, it is most likely that TiO₂ NPs possess a radiosensitising ability in reaction with X-rays, and this can be enhanced by conjugation with other materials to achieve significant dose enhancement effects. While X-ray-induced intracellular ROS appear to offer radiation therapy enhancement, other biochemical benefits of TiO₂ NPs are possible. As different types and concentrations of TiO₂ NPs with different cell lines have been reported so far, more comprehensive and comparable studies should be warranted to clarify the potential factors of TiO₂ NPs and their effective modifications, as well as the mechanisms underlying the dose enhancement for radiation therapy.