3.1.1 X-Ray Computed Tomography

X-ray Computed Tomography (CT) is among the most diffused and frequently employed imaging techniques in clinics. It is a quick and non-invasive diagnostic technique which allows for high resolution 3D imaging of large tissues, up to full-body. In CT imaging, the X-ray source and the detector rotate around the patient creating two-dimensional projections of X-rays attenuation through the body, which are then reconstructed by mathematical algorithms to produce a 3D rendering.

X-rays based imaging techniques, in general, rely on the attenuation of X-ray photons while passing through an object, depending linearly on its density and on the fourth power of the atomic number [1]. Photons are attenuated upon interacting with a medium by Rayleigh scattering, pair production, photoelectric effect and Compton scattering. However, photoelectric effect (Figure 3.1) is the dominant source of attenuation for the photon energies commonly used in X-ray diagnostics (30–150 keV).

Hence, we will briefly describe this physical phenomenon and its relevance in X-ray based diagnostics.

In the photoelectric effect, a photon impinging on a target is absorbed and produces the emission of an electron if the photon energy is equal or higher than the electron binding energy. In X-ray diagnostics, an X-ray photon can transfer its energy to an electron of the innermost shell of the target (named K), producing its ejection. Then, an electron from an outer shell replaces the vacancy created in the K-shell, following the emission of an X-ray photon having energy equal to the energy difference between the two shells involved in the process. The electron binding energy associated to the K-shell is called K-edge. It assumes great importance since it is characteristic of the material and, at this energy, X-ray attenuation shows a discontinuous trend due to the increased probability of photoelectric effect.

In X-ray diagnostics, it is easy to discriminate between tissues with highly different densities, such as bones and fat. On the other hand, the naturally occurring X-ray attenuation is not sufficient to produce appreciable contrast differentiation between diseased districts and healthy tissue, such as in case of tumors. Contrast agents are generally required in order to generate images with the highest contrast to the surrounding tissue and can: i) increase CT sensitivity and enhance differentiation among different tissues, ii) provide specific biochemical information of a tissue, or iii) enable evaluation of tissue/organ function or performance.

Hence, specific contrast agents comprising high atomic number materials with proper K-edge are needed to employ this imaging technique in cancer diagnostic. The most commonly used contrast agents for CT are both ionic and non-ionic forms of iodine-based compounds. These compounds are systemically injected, quickly distributed in the body, and finally excreted un-metabolized through hepatobiliary/renal pathways.

These traditional contrast agents show several limitations, among which: i) induced adverse effects as kidneys and thyroids dysfunctions, ii) allergic reactions, and iii) diagnostic limitations due to body clearance, which sometimes is too fast for prolonged imaging applications [3, 4]. On this regard, nanoparticle formulations of standard contrast agents have been proposed to increase blood residence and provide passive targeting accumulations in solid tumors. Among these, named blood pools, iodine-loaded liposomes have demonstrated prolonged blood circulation with respect to bare contrast media, providing high resolution images of mammary tumors in vivo in the rat model [5].

However, some drawbacks of iodinate agents are still retained even as nanosized formulation. In particular, the K-edge at 33 keV is highly unfavorable in many CT clinical exams, in which energies above 100 kVp are reached. Moreover, an effective diagnostic dose of iodinated contrast agent is usually in the molar concentration range. For example, an adult patient undergoing a selective coronary arteriography is injected intravenously with ~45 mL of Hexabrix (Mallinckrodt Imaging), a common clinically approved iodinated CT imaging agent solution, containing 24 g of ioxaglate (an equivalent to 14.4 g of iodine). To overcome these issues, alternative contrast agents based on inorganic NPs were investigated [6–9]. In particular, as shown in Figure 3.2, gold nanoparticles are an ideal radiopaque contrast media since gold has a high density, a high atomic number, and possess favorable K-edge at 80.7 keV [10]. For example, gold provides about 2.7 times greater contrast per unit weight than iodine. Furthermore, gold is inert, biocompatible, and is a straightforward material for the development of all-in-one nanoplatforms.

For example, Cole et al. reported ex vivo contrast-enhanced X-ray detection of breast microcalcifications through active targeting of hydroxyapatite employing bisphosphonate-modified gold NPs [11]. Zhang et al. achieved specific targeting of damaged bones with glutamic acid functionalized gold NPs, which accumulated in micro-cracks and resulting in a potential site-specific contrast agents for X-rays [12].

Direct comparison between commercial iodine-based contrast agents and gold NPs have been reported by many groups [13–15]. In particular, Sun et al. performed in vivo CT imaging of tumor bearing mice, observing a fivefold increase in the X-ray attenuation from the tumor 24h after tail injection of the NPs [16].

Interestingly, Dou et al. systematically investigated the relation between size of gold NPs in the range 3–50 nm and the CT contrast enhancement [17]. They observed that the attenuation didn’t increase linearly with NPs size. Indeed, 13 nm NPs showed X-ray attenuation comparable with the attenuation produced by particles bigger than 40 nm. The findings were also confirmed by Monte Carlo simulations. The authors speculated that this behavior is related to the distribution of gold atoms in the particles, which is responsible for secondary interactions, i.e. attenuation of characteristic X-rays produced by primary X-ray-induced photoelectric effect. This idea was confirmed by investigations on gold atoms distribution in NPs by varying the NPs size.

It’s worth mentioning that graphene oxide (GO) NPs have been very recently reported to serve as CT contrast agent when decorated by silver NPs. GO/Ag NPs provided 24h blood circulation time and afforded the in vivo contrast-enhanced CT imaging of lung, liver and kidneys. Notably, at the lowest dose (0.5 mg/kg of body weight), those nanoplatforms showed high specificity and sensitivity for the diagnosis of kidney dysfunctions, even 36 h after the injection [18].

3.1.2 Photoacoustic Imaging

Photoacoustic effect, discovered by Bell in 1880, is the generation of acoustic waves produced by thermal expansion of a piece of material, following absorption of photons. Briefly, when a medium is irradiated by a short-pulsed laser, part of the photon energy is dissipated as heat and temperature is increased locally, producing a raise of pressure. After cooling, pressure returns to its original value and this pressure waves, or sound waves, are detected by an ultrasound transducer.

Compared to ultrasonography, which is a sound in – sound out technique, Photoacoustic Imaging (PAI) is light in – sound out. PAI can be exploited by naturally occurring optical absorbers in the human tissues as endogenous contrast agents, such as hemoglobin and melanin. Several investigations on breast cancer patients are highlighted the great potential of PAI as clinical diagnostic tool (Figure 3.3) [19].

Commercially available low quantum yield dyes, such as methylene blue, indocyanine green and porphyrins, have been widely used as exogenous contrast media for PAI, since, upon light excitation, they produced enhanced local heating due to non-radiative decays. However, organic dyes have often low chemical stability, suffer from photo-bleaching, and are rapidly cleared from the body, being not suitable for prolonged PAI analyses. On this regard, nanomaterials, and in particular plasmonic metal NPs, have marked a major turning point in PAI technology, providing increased photostability, higher spatial resolution and selectivity.

Noble metal nanoparticles with anisotropic morphologies (nanoshells, nanocages, nanorods and branched NPs) are generally introduced as PAI contrast agents, due to their LSPRs in the NIR windows, allowing for deeper light penetration. In this regard, gold nanorods are ideal light-to-heat transducer. Indeed, they are among the most widely investigated nanomaterials for photoacoustic, mainly because of their tunable LSPRs in the NIR region and high extinction coefficient, which is almost all given by absorption. For example, Peng et al. have recently reported an in vivo PA investigation of both subcutaneous tumor and leg ischemia in mouse models by employing silica coated gold nanorods (Figure 3.4) [20].

Also, semiconducting nanoparticles, despite the absence of LSPRs, have demonstrated interesting behaviors as PAI probes. In particular, copper sulfide NPs have been employed as in vivo photoacoustic probes of brain, due to their strong absorption in the NIR region. Furthermore, copper sulfide NPs have allowed for ex vivo tissue imaging to up to 5 cm [21].

Organic nanomaterials have also been demonstrated to be suitable for PAI. For example, organic semiconducting polymeric NPs and dye functionalized carbon nanotubes have been reported to be excellent photoacoustic probes for in vivo cancer detection, exploiting their broad non-plasmonic absorption of visible to NIR light [22–24].

Agglomerates of self-organized small-molecules dyes, termed J-aggregates, were also shown to hold a good potential as PA probes with respect to single dyes, due to aggregation-induced self quenching and bathochromic shift [25, 26]. However, not so many paper have been reported on the in vivo use of those aggregates, due to their poor stability into complex biological media [27–29]. Improvements in conformational stability of dye aggregates was recently achieved by embedding them into lipid vesicles. Moreover, their employment in ex vivo PAI of head and neck cancer was reported [28].

Of particular interest, a novel paradigm for the production of exogenous photoacoustic contrast media has been recently proposed by Avigo et al. [30]. The Authors have reported the ex vivo PAI detection of hybrid nano-architectures composed by plasmonic gold nanoparticles and organic dyes. The nano-architectures were designed to exploit the fluorescence quenching effect of gold NPs to increase the non-radiative decay of dyes, increasing the photoacoustic signals and avoiding the NPs re-shaping induced by irradiation.

3.1.3 Positron Emission Tomography

Positron Emission Tomography (PET) is a molecular imaging technique in which the final image is produced by the detection of photons arising from the annihilation of a positron with an electron in the tissue.

When a beta plus (positron) emitter radiotracer is injected in the patient, it spreads in the body and, following radioactive decay, it emits a positron with a certain kinetic energy. The positron loses energy mainly by Coulomb scattering, and when it reaches thermal equilibrium, it annihilates with a tissue electron, emitting two high energy photons (511 KeV) back-to-back. These photons are detected in coincidence by two oppositely located detectors and the annihilation point is calculated by reconstruction algorithms with a spatial resolution of 1–4 mm [31].

The calculated image is the map of radiotracer activity in the body, which is indicative of physiological and/or pathological activities of the patient. On this regard, the turning point in the employment of PET was due to the synthesis of the radiopharmaceutical fluorodeoxyglucose (¹⁸F-FDG) [32]. ¹⁸F-FDG is a glucose analog, whose uptake is directly correlated to organs metabolic activity.

¹⁸F-FDG and many other glucose analogs have been also widely exploited in oncology for the diagnosis of a number of solid tumors. Indeed, many neoplasms are characterized by higher glucose consumption with respect to healthy tissues. However, PET contrast agents suffer from tissue non-specificity, poor accumulation and short blood circulation time, albeit several radiotracers based on physiological radioisotopes, such as ¹¹C, ¹³N, ¹⁵O and non-physiological ones like ⁶⁴Cu or ⁶⁸Ga have been synthesized. Moreover, spatial resolution is one of the most limiting features of this diagnostic tool, in particular on in vivo investigations.

These hurdles have been partially addressed with the rising of nano-radiopharmaceuticals, i.e. nanomaterials conjugated to radioisotopes which afforded the opportunity of: i) synergistic action between various imaging modalities, and ii) decoration by active targeting moieties.

Most of the radiolabeled nanomaterials employ chelating agents such as 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA), diethylenetriaminepentaacetic acid (DTPA), ethylenediaminetetraacetic acid (EDTA) or porphyrins, to load radioactive isotopes of metals, among which Cu, Y, Zr, In or Ga [34–36].

One of the most extensively used radioisotope, ⁶⁴Cu, has been conjugated both to organic NPs, as lipid vesicles and porphysomes, and to inorganic nanomaterials, such as gold NPs or silica NPs [37–42].

Unluckily, radiolabeling of nanomaterials by means of chelating agents has shown several drawbacks. In particular, the inherent instability of certain metal-chelator conjugates can result in losing the radioisotopes and subsequent non-specific binding to proteins [43]. Hence, an good knowledge in coordination chemistry is needed, in order to properly choose the most stable chelator for each radioisotope [34, 43].

Recently, new chelator-free strategies for radiolabeling of nanomaterials were developed. Several approaches for the production of intrinsically radiolabeled NPs have been summarized by Goel et al. They have identified four main synthetic strategies, among which an interesting post-synthetic labeling by proton beam induced nuclear reaction in metal oxides NPs [44]. Sun et al. reported a chelator-free radiolabeling route with ⁶⁴Cu on PEG stabilized gold NPs by directly reducing ⁶⁴Cu on gold NPs surface. They have also tested their feasibility as PET radiotracers in vivo [39]. Interestingly, Shi et al. have recently reported radiolabeling of nanographene by direct interaction between ⁶⁴Cu and π electrons following simple mixing of the two species in a proper buffer. This nanoplatform have demonstrated a good stability in vitro, in mouse serum, and was successfully employed in PET imaging in tumor-bearing mice [45].

Notably, a generalized route for post-radiolabeling ultrasmall iron oxide NPs was achieved by Boros et al., who reported a heat-induced chelate-free labeling with ⁸⁹Zr, ⁶⁴Cu and ¹¹¹In. They have also reported an in vivo investigation on NPs biodistribution by combined CT/PET imaging on mouse models [46].

3.1.4 Magnetic Resonance Imaging

Magnetic Resonance Imaging (MRI) is a diagnostic technique based on the physical principle of Nuclear Magnetic Resonance (NMR), discovered by Bloch and Purcell in 1946, and for which they were awarded the Nobel prize in physics.

In this section we will briefly describe the principles that regulates the MRI, without getting into the physical details of the NMR, which are comprehensively discussed elsewhere [47, 48].

When a static magnetic field B₀ is applied to a water-containing medium, the randomly oriented magnetic moments μ = γJ (where γ is the gyromagnetic ratio and J is the total angular momentum) of hydrogen nuclei, align along two anti parallel directions, being J = ±ħ/2, corresponding to two states¹ having energies E = –μΒ₀. B₀ exerts a torque on μ, causing a precession of J about B₀ axis with an angular frequency ω₀ = γΒ₀ called Larmor frequency.

At the basis of the MRI, there are two main points: (i) Larmor frequency is peculiar for each element or isotope, and (ii) it is possible to collect the signal of a specific element or isotope by exciting the system with a radio frequency (RF) magnetic field orthogonal to B₀, that matches its Larmor frequency. RF pulses cause the magnetic moments to flip from the low energy state, by an angle which depends on the pulse duration, revealing the magnetization of the system. When the RF pulse ends, the magnetization relax to its initial value, by producing itself an oscillating magnetic field, that is detected by a coil as a voltage signal, and is displayed as free-induction decay (FID).

The intrinsic time needed from the magnetization to return to its initial orientation after the end of the RF pulse is tissue-dependent, hence it is the source of contrast in MRI. In particular, two characteristic times are defined: (i) the transverse relaxation time, named T₂, and (ii) the longitudinal relaxation time, or T₁ (Figure 3.5).

T₁ is the time needed for the longitudinal component of magnetization to return to about 63% of its maximum value, i.e. spins lose energy by heat transfer to the environment, favoring the low energy orientations. In biological tissues, T₁ ranges from tenths of second to few seconds, a relatively long time, because the energy transfer requires stimulation by surrounding molecules which produces a field oscillating at a frequency close to the Larmor’s one.

T₂ is the time needed for the magnetization to decay to 37% of the value it possessed before the RF pulse, mainly through dephasing of ensembles of nuclear spins due to local magnetic field inhomogeneity.

The longitudinal and transverse relaxation times of protons in human tissues in both physiological and pathological states have been extensively investigated also as a function of temperature and magnetic field strength. From the tabulated reported by In 1986, Bottomley and co-workers have demonstrated that T₁ and T₂ values collected in neoplasms are usually quite increased respect to healthy tissues [50]. On the other hand, despite the increased water content of cancerous tissues, the differences of relaxation times due to necrosis, inflammations, or edemas respect to healthy tissues is negligible [51–54].

Hence, also for this imaging technique, contrast agents are needed to improve its efficiency for the detection of neoplasms.

The first nanoparticulate example of T₂ contrast agent was reported in 1986, when magnetite NPs at a concentration of 10 mg/kg body wt were injected intravenously in a dog model, inducing a net decrease in MRI contrast of liver and spleen [55]. Since nanosized iron oxide is superparamagnetic, it exhibits strong magnetization in presence of an external magnetic field, which disappears after the field is turned off. This magnetization creates local inhomogeneity of the magnetic fields, accelerating the nuclear spin coherence losing, thus shortening T₂. These tools are named negative contrast agent, because relaxation shortening produces a decrease in the signal intensity. Superparamagnetic Iron Oxide NPs (IONPs) are mainly exploited in detecting liver diseases, since IONPs are selectively internalized by Kupffer cells which are less numerous in tumor-bearing livers. Thus, after treatment with IONPs, damaged tissues appear brighter, i.e. with higher signal intensity, due to the poor uptake of negative contrast agents.

On this regard, Huang et al. investigated the T2 contrast enhancement in vivo depending on the IONPs size. They have performed MRI investigations with IONPs ranging from 8 to 65 nm on normal liver and hepatic lesions [56]. They found that 37 nm PVP-coated IONPs (HD: 100 nm) have the greatest contrast enhancement, in line with the maximum cellular uptake obtained with those NPs by macrophage cells (Figure 3.6).

Notably, Zhao et al. achieved effective T2 contrast enhancement and in vivo small liver tumors detection by employing anisotropic octapod IONPs. They reported stronger transverse relaxation of the octapods, in comparison with spherical IONPs of similar size, due to the strong magnetic field inhomogeneity induced by the anisotropic NPs under an external magnetic field, yielding stronger proton spin dephasing [57].

Many other groups have dedicated a lot of efforts to the investigation of IONPs as T2 negative contrast agents for diagnosis of breast cancer, brain glioma, transplanted stem cells tracking and, very recently, lung cancer [58–61].

The contrast agents which shorten the T₁, i.e. increase the MRI signal, are named positive contrast agents. Most of them are based on nanoparticles comprising chelating agents for Gd(III). Indeed, Gd ions cannot be directly administered in organisms because of the high toxicity due to strong interference with calcium ions channels.

Advances in the use of Gd-based positive contrast agents were related to the introduction of gadolinium oxide (Gd₂O₃) NPs, gadolinium fluoride (GdF₃) and Gd-Si oxide NPs, which were reported to offer a higher longitudinal proton relaxivity and, at the same time, an improved safety profile with respect to Gd(III) chelates [62–64]. Park et al. reported the in vivo MRI of brain tumor in mice model by employing ultrasmall Gd₂O₃ NPs coated by D-glucuronic acid [65]. These NPs have demonstrated an increased relaxivity and a significant reduction of toxicity respect to Gd(III) chelates.

Notably, Li et al. have recently reported the production of dual modal MRI contrast agents for both T₁ and T₂, by synthesizing Fe₃O₄/Gd₂O₃ core/shell nanocubes. Although other groups had previously reported the synthesis of dual modal MRI contrast agents, this was the first report of cubically-shaped NPs containing both iron and gadolinium [66–68]. Holding relaxivities about two folds higher than Fe₃O₄ nanocubes and Gd₂O₃ NPs, these hybrid nanostructures were exploited for MRI in murine models, showing high positive and negative contrast enhancement in the liver and producing negligible cytotoxicity and no induced morphological changes in both normal and tumor cell lines [69].

3.1.5 Raman-Based Diagnostics

Raman spectroscopy is an analytical technique that allows for fingerprint identification of chemical species constituting complex biological samples. Moreover, cells or tissues that are subjected to biochemical alterations due to diseases can be distinguished from healthy tissue by significant changes in the Raman spectra. Raman spectroscopy can provide fingerprints of tissues at molecular level, increasing the interest among scientists for its potential as promising diagnostic tool.

Raman spectroscopy relies on the collection of electromagnetic radiation triggered by a laser that is scattered either elastically or inelastically upon interaction with a material. The scattered light is collected and its intensity is plotted against: i) the frequency shift (Raman shift) with respect to incident light, or ii) the inverse of wavelength, named wave number, which is usually expressed in cm^-1 (Figure 3.7).

Most of the light is scattered elastically (Rayleigh scattering), whereas a photon every 10⁶ to 10⁸ exchanges energy with the medium through phonons mediated scattering processes [70]. These inelastically scattered photons give rise to the Stokes and anti-Stokes peaks in the Raman spectra, corresponding to light which have released energy to the analyte, or acquired energy from it, respectively.

Raman spectroscopy reveal the vibrational structure of the sample, since after having interacted inelastically with photons, the molecules are in a vibrational state which is different from the initial one, and the Raman shifts match exactly the energy difference between initial and final vibrational level of the analyte.

Raman peaks are narrow and can be associated to specific vibrations of chemical bonds [71]. Thus, Raman spectra depict a detailed chemical fingerprint of the molecular system under investigation, allowing for its rapid identification and providing sensitivity to conformational changes. Nevertheless, the detection of chemicals in complex biological media is limited by several drawbacks, such as the inherent low sensitivity of Raman technique towards electronically saturated systems, i.e. carbohydrates lipids or amino acids [72]. Indeed, Raman selection rules expects that the higher the polarizability of the analyte, the more intense are the Raman peaks, meaning that the best detectable species are molecules with extended π-electron systems, such as aromatic moieties.

A huge step forward towards increasing the sensitivity and selectivity of Raman spectroscopy was made after the experimental evidence that Raman spectra acquired with an excitation source whose frequency is closely matched an electronic transition of the analyte showed enhanced peak intensities [74]. This phenomenon, named resonance Raman effect, was exploited by anchoring chromophores to sites of interests, to achieve signal enhancement and high selectivity towards specific vibrational modes.

However, labeling biological samples of interests with small dyes to exploit the resonance enhancement have shown some drawbacks that prevented Raman spectroscopy to be employed as effective diagnostic tool. For example, the fluorescence background significantly reduces Raman signals collection. Indeed, typical fluorescence cross section of chromophores are in the range 10^–17–10^–16, while the respectively Raman cross sections are usually in the range 10^–30–10^–25cm² [75]. This issue is of particular importance in resonance Raman, since the laser source is energetically close to the electronic transition, increasing the probability of fluorescent emission, whose intensity and spectral broadening can easily overwhelm Raman peaks [76]. Moreover, the introduction of exogenous tags can potentially interfere with physiological conditions.

On this regard, metal nanoparticles have represented the turning point in the development of Raman-based highly sensitive diagnostics due to: i) the fluorescence quenching properties of some metals, and ii) the Raman signals enhancement [77–79].

Surface Enhanced Raman Scattering (SERS) pushed the inherently insensitive Raman spectroscopy to unprecedented levels of sensitivity thanks to the giant signal enhancement, up to 10¹⁵, that allow to single molecule detection [80]. The physical mechanisms at the basis of SERS are today well understood and arise both from the local electromagnetic field enhancement owing to the plasmonic effect (physical enhancement), and from an electronic coupling between the analyte and metal surface (chemical enhancement). The enhancement factors are strongly dependent on the SERS substrates, analytes and excitation wavelength; however, this subject is beyond the scope of this book, and it is comprehensively reviewed elsewhere [81].

Colloidal gold and silver NPs are among the most investigated SERS-active substrates for non-invasive early diagnosis of diseases, and many groups reported their use for the detection of clinically relevant analytes from biological fluids, such as cancer related proteins and DNA biomarkers [82–87].

The development of noble metal NPs whose plasmonic response is shifted up to near infrared, and the synthesis of novel NIR Raman-active molecules have allowed for the translation of SERS based diagnostics in vivo, owing to the greater penetration depth in tissues of NIR laser sources [88, 89]. Zhang et al. reported the in vivo SERS detection of tumors in a mouse model, by employing dye-labeled silica/polymer coated gold nanorods. Imaging was performed by mapping the intensity of one SERS peak (508 cm^–1) of 3,3’-diethylthiatricarbocyanine iodide (DTTC) which revealed an increased intensity at tumor sites due to accumulation of nanorods [90]. DTTC dye was also exploited as Raman reporter by many other groups [91, 92]. For instance, Conde et al. reported in vivo tumor inhibition and spectroscopic contour detection, i.e. theranostics, with a nanosystem composed by 90 nm gold NPs capped with DTTC as Raman reporter and an antibody-drug conjugate in mouse bearing xenograft tumors [93].

Tian et al. achieved in vivo real time SERS tracking of Mitoxantrone (MTX) chemotherapeutic drug release in a lung cancer mouse model, by employing NIR resonant gold nanostars as SERS probes. Notably, dye-labeling of gold nanostars was not needed, since drug release was imaged by evaluating the SERS intensity of aromatic C-C stretching of the MTX itself, which acted both as therapeutic and imaging tool [94].

Interestingly, Gandra et al. have recently synthesized gold/gold core/shell SERS nanoprobes with tunable interstitial gap that exhibited plasmon resonances at 785 nm, which showed exceptional stability in physiological conditions both in vitro and in vivo over 24 hours, allowing for prolonged non-invasive imaging of tumors in a mouse model (Figure 3.8) [95].

3.2.1 Chemotherapy

Chemotherapy, literally a therapy based on chemicals, is referred to the administration of cytotoxic chemical agents for the treatment of neoplasms. The first reported use of chemotherapeutic drugs for healing cancer dates back to 1940s, when nitrogen mustards were employed for the treatment of a patient lymphoma [96].

Chemotherapy is often used as adjuvant therapy after a surgical treatment, or as preoperative therapy with the aim of reducing the tumor volume before the surgery. In advanced stages, chemotherapy is used only as palliative care for providing relief from painful symptoms and slightly increase patients’ life expectancy [97].

The majority of drugs employed in chemotherapy interact with cells functions, altering their growth and division. A huge class of chemotherapeutic drugs, belonging from nitrogen mustard, is alkylating agents, which are chemicals that add an alkyl group to important biological macromolecules, including DNA, inhibiting their normal functions and causing double strand breaks, which eventually lead to cell death.

Another chemotherapeutic drug that inhibits DNA functions is cis-diamminedichloroplatinum (II) (cisplatin). It acts as DNA crosslinker, mainly causing intrastrand cross-link. Cisplatin attaches to DNA purine bases, preventing the opening of DNA double helices. Cisplatin was introduced in clinical investigations in 1971 and dominated the therapy of ovarian and testicular cancers for decades [98]. Many cisplatin analogues have been synthesized during the years, even though only cis diammine (1,1-cyclobutanecarboxylate) platinum (II), known as carboplatin, have received clinical approval owing to its reduced side effects and enhanced chemotherapeutic activity [99].

Another important class of chemotherapeutic drugs comprises antimetabolites, which include: i) cytarabine, employed for the treatment of acute myeloid leukemia, and ii) gemcitabine, a prodrug which is converted to its active form inside the cells upon kinase phosphorylation [100, 101]. These chemotherapeutics inhibit the nucleic acids synthesis. Antimetabolites include either: i) analogues of pyrimidines and purines, which replace the naturally occurring bases that constitute the building blocks in the synthesis of DNA and RNA impairing their production, or ii) antifolates, which are antagonists of folic acid, an essential vitamin in the production and repair of nucleic acids.

Topoisomerase inhibitors are a class of chemotherapeutic drugs that includes doxorubicin (DOX). DOX is a natural antibiotic extracted from soil fungi, whose cytotoxicity is produced by several mechanisms including: i) DNA intercalation, which inhibits the progression of the enzyme topoisomerase II, and ii) generation of cytotoxic free radicals, such as peroxides and superoxides [102].

Chemotherapeutics can induce cells death also by destabilizing microtubule polymerization, leading to their anomalous production or deficiency. Since microtubules play a pivotal role in a number of cellular activities, including proliferation, adhesion, migration and intracellular transport, an anomalous polymerization heavily influences the cell cycle, inducing apoptosis. One of the most successful chemotherapeutic drug acting on microtubules is paclitaxel, whose anticancer activity relies in the stabilization of microtubules reducing their dynamism and causing the block of mitotic phase by leading to apoptotic death [103].

Unluckily, none of these drugs is tumor-specific. Thus, upon administration, they spread all over the body causing severe adverse effects, in particular to fast proliferating cells: blood cells in the bone marrow, hair follicles, reproductive systems and tissues belonging to the digestive apparatus.

Hence, common side effects include hair loss, asthenia, headache, nausea and vomiting, severe blood disorders such as thrombocytopenia and leucopenia, diffused pain, cognitive dysfunctions including memory and attention loss or nerve damages. Moreover, some adverse effects may last for months/year after chemotherapy treatment, seriously prejudicing patient’s life quality and expectancy [104].

Overall, chemotherapeutic treatment is often doomed to failure due to the occurrence of drug resistance, either because of: i) inefficient drug distribution in tumor tissues, or ii) through molecular efflux pumps which provide extrusion of the drugs from the cells [105].

Therefore, increasing efforts are devoted to achieve localized and enhanced anticancer activity of drugs. Nanotechnology offer a concrete possibility for the development of targeted chemotherapeutics, owing to: i) passive accumulation of nanomaterials at tumor site through Enhanced Permeability and Retention effect (EPR, see chapter 2), and ii) combination of multiple drugs on the same nanoplatform.

During the last decades, a plethora of works were published on this subject, establishing nanomedicine as one of the most exciting field of academic research.

After the successful encapsulation of drugs in liposomes in 1973, one of the first significant clinical trials of chemotherapeutics nanoformulation was reported in 1996 by Nortfelth and co-workers [106, 107]. They proposed a liposome comprising doxorubicin against Kaposi’s sarcoma and reported an accumulation of liposome-encapsulated doxorubicin in the target to up to 11.4 times more respect to standard doxorubicin.

Triggered by this results, several liposomal formulations comprising various drugs, among which cisplatin and paclitaxel, have been produced, and, notably, there are 16 liposomal drug formulations that are currently clinically approved, among which six are treatments for cancer-related diseases [108–110]. Also metal NPs, in particular gold nanostructures have been extensively employed as drug carriers owing to their biocompatibility, the straightforward surface functionalization, and the possible on-demand drug-release by exogenous or endogenous stimuli [111, 112]. Owing to the high affinity of gold surface towards thiols, endogenously triggered release can be in situ achieved due to the difference between intracellular (1–10 mM) and extracellular (1–5 μΜ) glutathione (GSH) concentration. Indeed, GSH can act either as a reducing agent, cleaving moieties linked through disulfide bonds to gold surface, or as substituting coating agent, replacing the original surface capping layer [113].

Stimuli responsive gold NPs have been also engineered to be triggered by external laser irradiation [114–116]. On this regard, effective spatio-temporal control over release of cargo molecules was achieved through 1,2,3 triazolic systems functionalized 30 nm gold NPs, which allowed for externally triggered payload release inside target cells through synergic action of irradiation with 561 nm laser and local electromagnetic field enhancement induced by LSPR stimulation [117].

Polymer nanomaterials are other interesting candidates for nanoparticle-mediated chemotherapy. The most promising are PLGA NPs, due to sustained release properties, low toxicity and high encapsulation capability of hydrophobic moieties [118–122]. Very recently, Rodgers and collaborators reported the synthesis and in vivo application of PLGA NPs comprising both paclitaxel and cisplatin. They found that co-delivery of the two encapsulated drugs enhanced the chemotherapeutic efficacy against lung tumors relatively to both free drugs combination and also combined administration of single drug loaded NPs [123].

Owing to biocompatibility, easy designing, high drug loading and biodegradation to nontoxic byproducts, silica NPs hold a prominent position in promising chemotherapeutic development [124]. During the last year many groups have reported chemotherapy applications of silica NPs both in vitro and in vivo, also exploiting strategies for selective drug release, either by endogenous or exogenous stimuli [125–130].

In this context, Zhang and co-workers have recently reported a strategy for synthesizing biodegradable silica nanocarriers for gene/drug co-delivery, whose release is related to GSH-induced biodegradation [131]. In particular, they have employed a modified Stöber method for the synthesis of their nanoplatforms. DOX was mixed to tetraethyl orthosilicate (TEOS) and disulfide bond-bridged silane (BTOCD) in presence of ammonia in order to obtain drug-loaded silica nanoparticles in one step, which exhibited accelerated degradation kinetics owing to the instable disulfide bridges in the silica matrix. Upon proper surface functionalization with targeting moieties, these smart nanoplatforms were successfully exploited for in vivo therapy on glioma tumor bearing mice, inducing an interesting decrease in tumor size after the treatment compared free DOX (Figure 3.9).

A remarkable result was very recently reported by Dai et al.. They synthesized silica coated rare earth up-converting NPs able to emit photons at shorter wavelength than the absorbed ones, producing multidrug release upon irradiation with near infrared light source [132]. In their work, DOX was loaded into porous silica shell. Pores were closed by a photoactivatable moiety linked to a cisplatin prodrug. After irradiation with 808 nm laser, the up-converting NPs emitted UV photons that triggered the conversion of cisplatin prodrug into its active form, while simultaneously opening the gates for doxorubicin release. These multiphoton activated double-drug delivery nanoplatforms were exploited for in vivo chemotherapy on tumor bearing Kunming mice, reporting significant tumor shrinking while not producing any scars on the skin at the irradiation site, due to the short duration of laser pulses which prevent overheating.

3.2.2 Hyperthermia

HyperThermia (HT) is an ancient therapeutic procedure consisting in raise the temperature of the diseased body areas, or the whole body, to improve its healing. It is typically combined to other therapeutic techniques. HT is commonly defined as temperature increase up to 45 °C, as higher temperature treatments fall into the thermal ablation techniques [133].

For the treatment of cancer-related diseases, HT is either supplied alone or in support to chemo- and radiation therapy, enhancing their therapeutic action. HT, indeed, was proven to enhance the cytotoxic effect of many chemotherapeutic drugs, including the most widely used cisplatin, Doxorubicin, Paclitaxel and Gemcitabine, although the mechanisms underlying these enhancements are still not fully understood [134–137].

Interestingly, HT does not bear significant effect on cultured cancer cells viability, while induce a significant effect in vivo due to the vasculature of solid tumors, the highly acidic environments of neoplasms, and the usual marked hypoxia that cancer cells suffer [138].

HT treatment of solid tumors is locally supplied by either external electromagnetic or ultrasound sources or by heating the target with a hot fluid. Whole body HT is performed typically to treat diffused metastatic cancers, for which a large area needs to be covered.

The effectiveness of HT treatment against tumors has been ascribed to several temperature-dependent biological mechanisms, including impaired DNA double strand break repair, misregulation of gene expression, increased mitochondrial reactive oxygen species (ROS) production, cytoskeleton disruption resulting in altered intracellular transport, antitumor immune response activation and many others [139–144].

Despite the benefits brought by HT in oncology, there are several issues to be addressed to enable its widespread use in clinics. Indeed, the localization of the treatment is difficulty reached and not all the body’ districts are equally sensitive to temperature increase. For example, many tissues are unaffected by prolonged HT treatments (1 hour at 44 °C), while milder HT conditions (42 °C for 40 minutes) irreversibly damage the central nervous system [145]. Furthermore, heat exposure of large areas, up to whole body, results in thermo-tolerance development due to heat shock proteins (HSP) concentration increase, significantly reducing the benefits of repeated HT treatments [146].

Hence, also this technique needs an increased specificity of action to enhance the treatment effectiveness. One important step forward toward this goal was accomplished through the engineering of light-responsive NPs that produced heat upon interaction with non-ionizing radiation.

NPs assisted light-induced heat is mainly achieved by employing nanomaterials composed by low quantum yield dyes or plasmonic metal NPs or semiconductor quantum dots or carbon-based nanomaterials, each one relying on different physiochemical light-to-heat conversion mechanisms [147].

Gold nanorods (Figure 3.10) are among the most widely used nanomaterials for PT owing to: i) an exceptional conversion efficiency (namely the ratio between absorption and extinction cross sections), and ii) their finely tunable LSPR up to the NIR biological windows. One of the earliest example was published by Huang et al. in 2006 [148]. They reported a selective photothermal destruction of cultured cancer cells incubated with gold nanorods by irradiating at 800 nm in 40–200 mW power range, demonstrating an interesting dependence between nanomaterial uptake and cell apoptosis. A significant was achieved by Li et al., who perform in vitro photothermal investigations on gold nanorods-incubated HeLa cells by employing a circular polarized pulsed femtosecond laser beam [149].

Remarkably, Choi et al. developed smart chitosan-conjugated pluronic acid-based nanocarriers containing aggregates of gold nanorods, which were reported to yield in vivo higher photothermal efficiency relatively to single nanorods, achieving an apparently complete tumor resorption on bilateral tumor bearing mice [150].

In the last decade, several other investigations on in vivo effectiveness of gold nanorods for PTT have been reported [151–154]. In particular, very recently, Zhao and co-workers produced dual-stimuli responsive and reversibly pH-triggered gold nanorods for tumor-selective PTT [155]. Nanorods were functionalized by a pH-switchable asymmetric cyanine dye, which provide additional PT effect and fluorescence imaging, and a metalloproteinases targeting peptide, achieving PTT in mice bearing subcutaneous tumors (Figure 3.11).

Other gold nanostructures employed for both in vitro and in vivo PTT include nanoshells, branched NPs, nanocages, nanocubes and the unusual gold/silica geometry named nanomatryoshkas [156–162]. The latter has shown higher PT transduction efficiency in a direct in vivo comparative study against gold nanoshells.

Furthermore, PTT was obtained also by employing other plasmonic metal nanomaterials (such as Pt, Pd and, very recently, titanium nitride NPs), carbon-based nanomaterials (including carbon nanotubes, graphene and graphene oxide NPs), semiconductor and polymeric NPs [163–176].

Another class of external-triggered heat-generating nanomaterials is represented by magnetic NPs. Magnetic NPs can produce heat under the action of an alternating magnetic field. The heating efficiency is proportional to the frequency of the magnetic field. The physical principles at the basis of heat generation by magnetic nanomaterials are related to the NPs size and the strength of the applied magnetic field, and can be summarized into three mechanisms: (i) eddy currents, (iii) hysteresis loss and (iii) Neel and Brownian relaxation.

Eddy currents have usually a low contribute to magnetic NPs enhanced HT because the NPs small size and their low electrical conductivity [177, 178]. In turn, heat generation by hysteresis losses is ascribed merely to thermodynamic conservation of energy: when an external magnetization is raised from M₀ to a certain value M₁ by an external magnetic field H, if the reversal of H does not bring the magnetization back to M₀, i.e. in case of hysteresis, the nonzero difference of internal magnetic energy between initial and final state equals the work made by H plus heat. Hysteresis losses occur both in multi-domain materials, due to domain walls reorganization, and in single-domain ones, according to Stoner and Wohlfarth, when magnetic anisotropy is so high that coherent reversal of atomic magnetic moments is inhibited [179].

Heating by single-domain NPs occurs also through magnetic moment orientation fluctuations either due to Brownian physical motion of the NPs in the fluid or owing to fluctuations of the magnetic moment within the particles (Neel fluctuations). Upon application of an external alternating magnetic field, both the phenomena occur with different relaxation times, though efficient thermal transduction of magnetic energy is achieved only when Neel relaxation is not the dominant process [180].

The first report of magnetic NPs assisted HT goes back to 1957, when IONPs of size varying between 20 and 100 nm exposed to a radio frequency alternating magnetic field were shown to produce heat-induced necrosis in lymph nodes at the concentration of 5 mg NPs per gram of tissue [181]. Since then, the physical mechanisms at the basis of heat generation by magnetic NPs have been deeply investigated, with a particular focus on understanding collective effects produced by interparticle (e.g. dipolar) interactions, of undoubted relevance for practical cases [182–184]. Indeed, Dennis and co-workers provided the experimental evidences of the enhanced heat generation produced by collective interactions of magnetite NPs, which were successfully employed for in vivo HT, leading to almost complete regression of mammary tumor in mouse model [185].

Recent efforts in addressing the challenge of controlling interparticle interactions to achieve enhanced HT efficiency were published by Andreu et al., who reported the synthesis of 1D worm-like and 2D spherical surface arrangements of magnetite NPs, employing silica and PLGA nanocarriers, respectively [186]. They have demonstrated that the geometrical arrangement is the key-feature for the thermal generation efficiency, rather than the concentration of magnetite NPs.

With the aim of combining more potential therapies in a single nanoplatform, synergistic action of PT and magnetic assisted HT was achieved by synthesizing NPs bearing both plasmonic and magnetic features. It was demonstrated that upon exposure of these hybrid magnetoplasmonic NPs to laser irradiation and radiofrequency magnetic field, the cumulative effect of bimodal PT and HT led to a remarkable local increase of temperature in vivo, reaching 48 °C at the tumor site in a mouse model (Figure 3.12) [187].

Ding et al. demonstrated the enhanced therapeutic effect produced by hybrid silver-IONPs upon HT treatment. They have observed that the presence of silver together with magnetic NPs, either in a core shell or dimeric arrangement, led to an increased tumor suppression, which the authors attributed to the synergistic effect of magnetic HT and silver ions release triggered by the increase in temperature [188]. Several groups have also reported the synthesis of complex nanoplatforms comprising both drugs and magnetic NPs, that have the potential for combined HT/chemotherapy or for HT-triggered drug release [189–192].

A plethora of other in vivo investigations on NPs assisted magnetic hyperthermia have been introduced during the last years, confirming the excitement of the scientific community towards this promising non-invasive treatment [193–196].

3.2.3 Radiotherapy

Along with chemotherapy, radiation therapy (or radiotherapy, RT) is one of the most employed cancer treatment. Indeed, about 60% of oncological patients experienced RT sessions, either as primary or adjuvant therapy. In RT, high doses of ionizing radiations (X-rays, γ-rays, hadrons) are administered either with external beams or with radiation sources injected in the body (brachytherapy), to shrink tumors.

The effect of RT is related to either directly damaging fundamental biological macromolecules, or by producing free radicals, such as reactive oxygen species (ROS). Even though ionizing radiations induce damage indiscriminately to healthy and diseased tissues, cancer cells are often unable to recover from DNA damage owing to inefficient repair pathways [197]. Hence, RT is administered in multiple sessions at low doses, repeated after a rest period which allows only healthy tissues to recover from the inflicted non-lethal damage.

Nonetheless, RT involving low radiation doses is often inefficient against cancer, since many tumors display an inherent resistance to ionizing radiations, owing to their hypoxic environment which severely limits ROS-induced cytotoxicity. On the other hand, an aggressive treatment plan could effectively damage tumors irreversibly while, in turn, increasing the likelihood of off-target injuries.

Hence, increase and localize the radiation dose at the tumor site by employing radiosensitizers is one of the most sparkling field of investigation. On this way, inorganic NPs of high atomic number elements are particularly suitable for this task, owing to their increased cross section for photoelectric effect [198].

Kobayashi et al. have reviewed the mechanisms at the basis of radiation dose enhancement by high-Z elements, stressing the role of Auger effect (a secondary low energy electron cascade emission that ensue photoionization) as the main responsible of local increase of dose deposition [199]. Also, although not explicitly referred to the Auger cascade, early experiments on increased radicals production in close proximity of 2 nm gold NPs irradiated by X-rays, have demonstrated the release of secondary electrons [200].

Gold NPs displayed interesting radiosensitization features even at radiation energies in the megavolt range, at which the photoelectric effect becomes negligible [201]. This feature has raised reasonable questions regarding the physical bases of radiosensitization of gold NPs, since it cannot be completely addressed to the high photon absorption cross section of gold nor to gold-induced oxidative stress [202]. Recently, it has been reported that the radiosensitization ability of gold NPs can be potentially influenced by their shape [203]. Indeed, it was observed on living cells experiments that gold nanospheres, nanorods and nanospikes of similar sizes produces different radiosensitization enhancements upon X-rays irradiation, with the highest enhancement reported for spherical gold NPs. Despite that, it is worth to notice that when the sensitization enhancement was normalized by the cellular uptake the differences become negligible, revealing that the increase in radiobiological effects is solely dependent by the increased NPs cellular internalization.

Notwithstanding incomplete knowledge of the radiosensitization mechanisms, gold NPs are a very promising and the most investigated nanomaterials for enhanced RT. Indeed, in the last years, plenty of works report both in vitro and in vivo investigations [204–210]. Nonetheless, other materials have been explored for enhanced RT, including cerium oxide, hafnium oxide, bismuth sulfide, gadolinium and silver [211–220]. In particular, Liu and co-workers investigated the radiosensitizing properties of silver NPs by a direct benchmark against gold NPs of the same size (15 nm), surface coating and concentration, upon irradiation by X-rays at megavolt energies [221]. Radiosensitizing efficacy of both NPs was tested in vivo on glioma-bearing mice after intratumoral administration, finding that Ag NPs yielded significant increase in antiglioma activity when irradiated at 6 MV X-rays, compared to their Au counterparts. The authors suggest that this enhanced radiosensitization could be ascribed to the increased level of autophagy induced by silver NPs with respect to gold.

However, several studies on the cytotoxic effects induced by silver nanostructures indicate that the radiosensitization properties of Ag NPs are likely a consequence of the increased ROS generation ensuing the release of Ag+ ions, rather than a high-Z effect [222, 223]. Also iron oxide NPs have shown interesting radiosensitizing effects, which were attributed to the release of Fe³⁺ ions which enhance the production of ROS upon X-ray irradiation [224–226].

Enhanced RT was explored for the local treatment of solid tumors, but in presence of metastatic diseases brachytherapy is preferred. Brachytherapy treatment relies on to systemic injection of radiopharmaceuticals (typically Auger electrons or α-and β-particles) [227–229]. By this approach, ionizing radiation sources wander around the body, hence, achieving tumor specificity is crucial in order to prevent radiotoxicity in healthy organs.

On this regard, several nanocarriers for radioisotopes delivery have been developed, including organic and inorganic nanoformulations of Iodine-131, Yttrium-90 and Rhenium-88 [230–237]. For instance, Tian and co-workers have recently reported the synthesis and in vivo investigations of ¹³¹I-tagged human serum albumin (¹³¹I-HAS) bound to manganese oxide NPs (¹³¹I-HAS-MnO₂). Notably, in vitro viability assays on 4T1 cells confirmed that ¹³¹I-HAS-MnO₂ significantly decrease cells viability, proving that cytotoxicity was induced by accumulation of Iodine-131. Furthermore, in vivo brachytherapy was accomplished through tail administration of ¹³¹I-HAS-MnO₂, yielding a preferential accumulation at tumor site compared to free ¹³¹I-HAS, significantly reducing the tumor growth (Figure 3.13) [238].