Magnetic targeting with superparamagnetic iron oxide nanoparticles for in vivo glioma

2.1 Search strategy

We included studies published between June 1999 and October 2016 that were found in the following databases: Cochrane Library, PubMed, SCOPUS, and Web of Science. A Boolean strategy was applied. The following sequence of keywords and Boolean operators was used: (DecS/MeSH): [(iron oxide) OR nanoparticle OR SPIO OR SPION] AND [(magnetic field) OR (magnetic targeting) OR magnetofection NOT AMF NOT (alternating magnetic field) NOT (AC magnetic field)] AND [(brain tumor) OR glioma OR glioblastoma OR gbm OR (giant cell glioblastoma) OR gliosarcoma OR astrocytoma OR (gliomatosis ependymoma) OR ependymoma OR oligodendroglioma OR oligoastrocytoma OR astroblastoma OR gangliocytoma OR ganglioma].

The classification of tumors of the central nervous system, including all glial tumors, standardized by the World Health Organization [8], was sought.

2.2 Data extraction

In this review, seven of the authors independently applied the search strategy and randomly selected reports for each disease in the databases cited above.

MFA, JBM, and LFG independently searched the studies. Discrepancies in the selection of studies between the two reviewers were discussed with a third reviewer, and concensus was reached. The analysis process and table plots of this study were carried out by full consensus of coauthors, respecting the distribution above. In cases of disagreement, a third independent author resolved the differences by data addition or subtraction. Next, JBM and LFG reviewed the nanoparticle characteristics, MFA and RFR reviewed the in vitro characteristics, MFA, MPNS, LPN, and HRS reviewed the animal model characteristics or in vivo assays, and MFA, TKF, JBM, MPNS, and LFG reviewed magnetic targeting methods used to assess therapeutic efficacy. In cases of disagreement, a third independent and senior author decided by data addition or subtraction.

2.3 Selection criteria

This review included only original articles written in the English language that reported on in vivo magnetic targeting for SPION targeting in gliomas or on the mediation of SPION targeting on the influence of an external magnetic field. Publications indexed in more than one database (duplicate mentions), incomplete data, conference presentation, book chapters, and reports written in a language other than English, as well as studies that did not use SPION, were excluded from this review.

3.1 Physicochemical characteristics of SPION

The physicochemical characteristics of SPIONs used in selected studies are summarized in Table 1. Of the 23 selected studies, 10 studies used SPIONs synthesized in the laboratory (“in lab”) [37], [38], [40], [41], [42], [45], [46], [49], [50], [55], nine used commercial SPIONs [39], [43], [44], [47], [51], [53], [57], [58], [59], and four studies looked at commercial and synthesized SPIONs [48], [52], [54], [56]. Most publications reporting on commercial SPIONs used FluidMAG [43], [47], [48], [51], [52], [53], [56], [57], [58] (Chemicell, Berlin, Germany), all used the FluidMAG-D product, only one study [51] used FluidMAG-ARA, two studies [52], [56] used FluidMAG-CMX, and one used FluidMAG-Heparin and FluidMAG-DEAE [56]. The other commercial SPIONs used were Magnetic Fluid-Amine [39], Advanced Nanotech [44], Resovist [54], and Ferrofluidics [59].

Eight studies [42], [45], [48], [49], [51], [52], [54], [56] used more than one type of SPIONs for a comparison criterion; four studies [48], [52], [54], [56] compared commercial products with synthesized in lab, and four studies compared only SPION synthesized [42], [45], [49], [50]. One study compared only commercial SPIONs [51]. The other studies did not compare the effect of the different SPION types and used only one type in the study, either commercial [39], [43], [44], [47], [53], [57], [58] or synthesized [37], [38], [40], [41], [46], [55]. It is noteworthy that the Pulfer study [59] compared the EMG-111 commercial nanoparticle with microspheres containing SPIONs. However, because the particle size was 1000–2000 nm, this study was not included for this review, in accordance with the previously established selection criteria.

In 14 studies [37], [38], [40], [41], [42], [45], [46], [48], [49], [50], [52], [54], [55], [56], the nanoparticles were synthesized by several chemical methods for SPION production in magnetic targeting. Among the chemical methods used were microemulsions [40], single-phase reaction (one-pot synthesis) [42], [45], hydrothermal reaction [46], sonochemical reaction [52], [55], and the coprecipitation method [49], [50], [56]. The SPION core chemical process for coating particles and conjugations with various materials to produce biocompatible particle was identified in only 14 studies [37], [38], [40], [41], [42], [45], [46], [48], [49], [50], [52], [54], [55], [56], and the crystal phase of iron oxide was identified in 12 studies [37], [38], [40], [42], [45], [48], [49], [50], [52], [54], [55], [56], 11 studies used a magnetite (Fe₃O₄), one study identified the crystalline phase as maghemite [37]; and two studies [41], [46] did not indicate which of the iron oxide crystalline phase was obtained.

The main coating material for the commercial SPIONs used was starch [43], [47], [48], [51], [52], [53], [56], [57], [58]. Other materials were dextran [39], DOX@PEG/PEI/Ps80 (doxorubicin/poly-ethylene glycol/polyethylenimine/Polysorbate 80) [38], carboxymethyldextran (CMD) [52], [56], carboxydextran [54], heparin, and dextran-diethylaminoethyl [56]. Among the studies summarizing the SPIONs used, the coating materials included DNPH (starch cross-linked, aminated, poly(ethylene glycol) (PEG)ylated and heparin-conjugated magnetic nanoparticles) [42], BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea) [45], Dextrana-Fluorophore-arginine-glycine-aspartate (Dx-F-RGD) [46], SPAnNa-BCNU (poly[aniline-co-sodiumN-(1-one-butyricacid) aniline]-BCNU) [49], SPAnH(poly-[aniline-co-N-(1-one-butyric acid) aniline]) [50], SPAnH-epirubicin [54], and gum arabic [56]. One study [37] used SPION complex to make rhodamine-labeled magnetic fluid-loaded PEG-ylated liposome (MFL) nanoparticle. Some studies have analyzed commercial SPIONs [43], [44], [51], [59] or involved synthesized SPIONs that used a combination of materials to form biocompatible coverage. Fan [39] used commercial SPION coating with dextran to develop SPIO-doxorubicin-conjugated microbubbles. Pulfer [59] used dextran particles with aminodextran and coverage, but as mentioned above, these are treated microspheres and were not considered in this review.

The physicochemical characteristics of the commercial SPIONs used were (I) Magnetic Fluid-Amine (MagQu Co Ltd, New Taipei City, Taiwan) with a magnetite core, average hydrodynamic size of 35.7 nm, coated with dextran, and provided at an iron concentration of 8.12 mg/ml; (II) Fl1uidMAG-d (Chemicell^®, Berlin, Germany) with a magnetite core, average core diameter of 12 nm, average hydrodynamic size of 50–251 nm, coated with starch, and provided at an iron concentration of 25 mg/ml; (III) FluidMAG-CMX (Chemicell^®, Berlin, Germany) with a magnetite core, an average hydrodynamic diameter of 100 nm, covered with carboxymethyldextran, and provided at an iron concentration of 25 mg/ml; (IV) FluidMAG-ARA (Chemicell^®, Berlin, Germany) with a magnetite core, average core diameter of 10 nm and 225 nm average hydrodynamic size, covered and provided with glucuronic acid at an iron concentration of 25 mg/ml; (V) Heparin-FluidMAG (Chemicell^®, Berlin, Germany) with a magnetite core, average hydrodynamic size of 100 nm, coated with heparin, and provided at an iron concentration of 10 mg/ml; (VI) FluidMAG DEAE (Chemicell^®, Berlin, Germany) with a magnetite core, average hydrodynamic size of 100 nm, coated with diethylaminoethyl-dextran, and provided at an iron concentration of 25 mg/ml; (VII) Resovist^® (Bayer Schering Pharma AG, Berlin, Germany), which are nanoparticles of a crystalline phase mixture of magnetite and magnetite, average core diameter of 5.9 nm and 63.8 nm average hydrodynamic size, coated with carboxydextran, and provided at an iron concentration of 27.9 mg/ml; (VIII) Advanced Nanotech (Taiwan), with a core diameter of 6–10 nm and 700–4000 nm hydrodynamic size, and provided at an iron concentration of 0.68 mg/ml; (IX) EMG-111 (Ferrofluidics, Corp. Nashua, NH, USA) with an average core diameter of 10 nm and 20 nm average hydrodynamic size.

3.2 Characteristics of cells used for tumor induction

The characteristics of cells used for tumor induction for further in vivo studies are summarized in Table 2. Most of the selected studies used glial cells of origin [41], [42], [43], [44], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], with only four studies [37], [40], [45], [59] specifically using cells that originated from a glioblastoma rat, called RG21. Among all the studies, only that by Fu [46] and Marie [37] used human cells, designated U87MG, and capable of originating a grade IV astrocytoma or glioblastoma. All other studies used cells from Rattus norvegicus [38], [39], [40], [41], [42], [43], [44], [45], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59] species. Of these, the most widely used cell line was 9L, identified in 11 studies [41], [42], [43], [47], [48], [51], [52], [53], [56], [57], [58].

The 9L cells originate the gliosarcoma tumor (more aggressive) in rats, mimicking important characteristics of human tumors, such as the pattern of tumor growth, growth factor expression, and metastasis. Another cell line widely used was the C6 (more solid and circumscribed with rat tumor), reported in seven studies [38], [39], [44], [49], [50], [54], [55]. These cells were cloned rat glioma induced by N-nitrosomethylurea [60] and have been widely used for a variety of studies on tumor growth, invasion, migration, neovascularization, growth factor regulation, and biochemical studies because they are morphologically very similar to glioblastoma multiforme [61], [62].

3.3 In vivo experimental model of glioma

The experimental model characteristics, as well as tumor induction parameters, are described in Table 3. All studies selected for this review are in preclinical testing and have used experimental models of rats or mice and clinical trials that have been performed. The small number of clinical studies on this subject shows that progress is early, and more work is needed on the nanobiotechnological tools for the diagnosis and treatment of gliomas.

Among the rodents identified, seven studies used mice [37], [40], [41], [42], [43], [45], [46], 17 used rats in their experiments [38], [39], [44], [45], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59], and Huang used both rats and mice [45]. From the studies that were based on rat models, the most commonly used species was Fisher 344, identified in nine studies [47], [48], [51], [52], [53], [56], [57], [58], [59]. The second most used species were the Sprague-Dawley rats, reported in seven studies [38], [39], [44], [49], [50], [54], [55]. Only one study [45] used Wistar rats. With regard to studies that used mice, the main species was nude mice, identified in four studies [37], [40], [42], [43]. Only one study [46] used SCID mice, and another study [41] used BLACK C57BL6 species.

The weights of the animals used in the various studies varied widely. For studies that used Fisher 344 rats, the weights ranged from 125 to 150 g in seven studies [47], [48], [51], [52], [53], [57], [58], 200 g in the study by Zhang [56], and ranged from 225 to 250 g in the study by Pulfer [59]. Among the studies that used male Sprague-Dawley rats, the weights varied between 200 and 250 g in Fan’s study [44], 250 and 300 g in Hua’s study [49], 320 and 350 g in the study by Chen [50], and 300 and 400 g in other studies [54], [55]. Huang [45] used Wistar rats and mice, both nude, in his study but did not provide the selected weight. Of the other three studies that used nude mice, the weight ranged from 18 to 22 g in Zhang’s study [42] but was not reported in the studies of Fang [40] and Zhou [43]. Only one study [41] that used C57BL6 black mice selected animals weighing 20–25 g, and Fu [46], the only study using SCID mice, considered animals weighing 28 g.

Regarding the sex of the rodents used, three studies [37], [40], [45] used females, while most selected male rats or mice [38], [39], [42], [44], [46], [47], [48], [49], [50], [51], [52], [53], [55], [57], [58], [59] and four studies [41], [43], [54], [56] did not provide this information. A few studies have reported the total number of animals and number of animals used per group. Among the studies that did provide this information, the variability was wide, ranging from three to eight animals per group [38], [39], [40], [45], [48], [49], [50], [53], [56], [57], [58], [59].

The choice of animal model is often directly related to the selected cell line. Thus, most studies that have opted for cell line 9L inoculated these cells in Fisher 344 rats [47], [48], [51], [52], [53], [56], [57], [58]. However, two studies inoculated 9L cells in nude mice [42], [43], and one study did so in black C57BL6 mice [41]. However, the study by Pulfer [59] used the cell line RG2 in Fisher 344 rats, unlike Fang [40] and Huang [45], who used this same lineage but in BALB/c nude mice and Wistar rats, respectively. All authors who used the C6 cell line chose the Sprague-Dawley [38], [39], [44], [49], [50], [54], [55] species. The only study that used a human cell line (RG2) inoculated these cells in SCID mice [46], which are useful as models for the lack of immunity of B and T cells and, hence, are susceptible to human disease in a chronic state.

The number of inoculated cells for tumor induction also differed in relation to both the cell chosen as the site of the tumor and the species used. Studies induced subcutaneous tumor inoculated [40], [41], [42], [43], [44], [45], [46], [47], in general, proportionally inoculated more cells. Among the studies that implanted tumor cells in the mouse flank, two used 2×10⁶ 9L cells [41], [43], two studies used 5×10⁷ and 10⁴–10⁵ RG2 cells [40], [45], one inoculated 10⁹ 9L cells [42], one study used 2×10⁷ U87MG cells [46]. Another site frequently used for tumor induction was the right cerebral hemisphere, identified in 11 studies [37], [39], [47], [48], [51], [52], [53], [56], [57], [58], [59]. In this place of glioma induction, the studies that used rats as an experimental model inoculated 10⁶ 9L cells [47], [51], [52], [53], [57], [58], 10⁵ 9L cells [46], [56], and 2×10³ RG2 cells [59]. The striatum was also used as a place of tumor induction and was chosen only for the rat model whose tumors were induced between 10⁴ and 10⁶ C6 cells [37], [38], [39], [49], [50], [55]. Fan [44] inoculated 2.5×10³ 9L cells but did not report the site of induction. Liu [42] also did not provide the site of tumor induction but reported the use of 10⁶ C6 cells to induce the tumor. Therefore, the major model used in the selected studies was lineage 10⁶ 9L cells inoculated into the right cerebral hemisphere (striatum) of male Fisher 344 rats weighing 125–150 g.

The vehicle suspension for inoculation of the tumor cells in animals was mostly DMEM fetal bovine serum free, identified in 11 studies [41], [42], [43], [47], [48], [51], [52], [53], [56], [57], [58]. Pulfer’s study [59] also used DMEM as a vehicle but did not specify whether it was fetal bovine serum free. Other vehicle suspensions were PBS [38], [39], [40] and saline [45] and MEM [49], [50], [55], while three studies [44], [46], [54] did not report the vehicle of choice. A few studies reported the duration of administration of the cells to the tumor site. When mentioned, the duration was 10 min in nine studies [38], [39], [44], [49], [50], [52], [55], [57], [58] and 30 s in one study [59].

The coordinates used for tumor induction also varied among the studies. In studies that inoculated cells subcutaneously, this information does not apply; however, the studies that induced glioma into the right hemisphere of the brain or into the striatum region were based on depth coordinates that were latero-lateral and anterior-posterior in relation to the stereotactic atlas. The average depth was 4.5 mm in five studies [44], [49], [50], [55], [59] and 3 mm in the other 10 studies [38], [39], [47], [48], [51], [52], [53], [56], [57], [58]. Only Liu [54] and Marie [39] did not report the depth used for inoculation of the tumor cells. In relation to the anteroposterior coordinate (axis), some studies reported that the cells were inoculated in the right hemisphere at 1 mm anterior to the bregma [48], [52], [53], [57] or 2 mm posterior to the bregma [59]; others did not provide this information [47], [51], [56], [58]. The latero-lateral coordinate used in the studies that deployed the cells in the right hemisphere was 5 mm [48], [52], [53], [57], 2 mm [59], or not reported [47], [51], [56], [58]. Among the studies [37], [38], [39], [49], [50], [55] whose administration of the cells was into the striatum site, the anteroposterior coordinate was 0.5 mm and 3 mm latero-lateral in relation to the bregma.

3.4 Experimental design of magnetic targeting

The parameters of the transfection process mediated by nanoparticles, also called magnetic targeting, are described in Table 4. For this review, we selected studies that performed magnetic targeting by imposing an external static magnetic field for more efficient and specific targeting of SPION at the tumor region.

The magnetic targeting process consists in SPION systemic injection for later magnetic targeting. Most studies administered SPIONs on the 10th day after tumor induction [44], [46], [47], [49], [51], [52], [53], [54], [55], [56], [57], [58] (described in Table 3). Two other studies administered SPIONs 7–14 days after tumor induction [40], [45]. Pulfer’s study [59] performed the procedure 14–21 days after induction. The other studies [41], [42], [43] did not specify this information. However, the appropriate day to perform magnetic targeting process was reported as that on which the tumor volume (mm³) is 300–400 [41], 300–500 [42], or 100–200 [43]. Even among articles that determine the day of SPION administration in animals, most take into account tumor volume [40], [44], [47], [48], [51], [52], [53], [57], [58].

The concentration of the nanoparticles (mg Fe/kg body weight) administered in the studies varied. The concentration used most often was 12 mg Fe/kg [41], [43], [45], [47], [48], [51], [52], [56], [57], [58]. Hua [49] used a concentration of 10.5 mg Fe/kg and Huang [55] used 10 mg Fe/kg, while Pulfer [59] administered nanoparticles at a concentration of 4 mg Fe/kg. Zhang [42], Chen [50], and Chertok [53] used different concentrations of nanoparticles ranging from 4 to 20 mg Fe/kg, 0.5 to 8 mg Fe/kg, and 12 to 25 mg Fe/kg, respectively. Marie’s [37] study had the highest concentration of iron (29 mg Fe/kg). The other studies [40], [44], [46], [54] did not report this information. Regarding the suspension vehicle used to administer the SPION in different routes, most of the studies reporting this information used saline [39], [40], [44], [50] or PBS [37], [38], [47], [48], [51], [53], [56], [57], [58]; Pulfer [59] used Tween 80/saline suspension as a vehicle. The most frequent administration route was intravenously, identified in 17 studies [37], [38], [39], [40], [41], [42], [43], [45], [46], [47], [48], [51], [52], [53], [56], [57], [58]. Other studies administered nanoparticles via the jugular vein [44], [49], [50], [54], [55] or carotid artery [47], [51], [59]. Of these, Chertok [47], [51] used tail injection and carotid administration in order to compare the efficiency of magnetic targeting process performed in different routes. Only three studies reported the duration of nanoparticle infusion: 10 min in two studies [57], [58] and 2 min in the study by Pulfer [59]. The main route of administration of SPIONs in selected studies was tail vein, with 12 mg Fe/kg in PBS vehicle.

As is known about the in vivo magneto targeting process, all studies used an external and static magnetic field with different characteristics for directing nanoparticles to the tumor region. Regarding the intensity of the field used, four studies [40], [46], [48], [54] used a magnetic field of 0.2 T, nine studies [38], [41], [42], [43], [47], [49], [51], [55] used a magnetic field of 0.3–0.35 T, and five studies [37], [39], [52], [53], [57], [58] used only 0.4–0.5 T. The highest intensity values found were 0.6 T [59] and 1.18 T [45]. Chertok [47] and Liu [54] used different field strengths to promote magnetic targeting: from 0.15 to 0.35 T and from 0.2 to 0.55 T, respectively. Only two studies [44], [56] did not report the intensity of the external magnetic field used. The duration of application of the magnetic field in each study also varied: 0.25–1.5 h [42], 0.08–48 h [45], 0.16–2 h [46], 3–24 h [54], and 2–24 h [55]. Most other studies applied the magnetic field for only 30 min [47], [51], [52], [53], [56], [57], [58], [59], four studies [38], [41], [43], [48] applied it for 1 h, Fang [40] did so for 6 h, and Hua [49] and Chen [50] did so for over 24 h. External and static magneto were used in most studies, by using a magnetic field of 0.3 T and 30 min duration.

To optimize the efficiency of the magnetic field, many studies used adaptations to the field for better targeting. Seven studies [44], [52], [53], [56], [57], [58], [59] did not complement the magnet with any artifice. Among the adjustments used, Fang [40] affixed the magnet to the animal’s head using a bandage, Zhou [41], [43] used a cylindrical neodymium magnet connected to the main magnet, Zhang [42] linked one small cylindrical magnet 6.4 mm in diameter to the other third cylindrical neodymium magnet 51 mm in diameter, Huang [45] used a directed magnetic field 1 cm in diameter, Fu [46] chose a rectangular neodymium magnet with dimensions of 22×7.5×7.5 mm³, Chertok [47], [51] used an electromagnet coupled to a cylindrical neodymium magnet 9 mm in diameter, Cole [48] used a cylindrical neodymium magnet 9 mm in diameter secured to a pole 40 mm in diameter, Hua [49], Chen [50], and Huang [55] set the magnet in the skull of rats, and Liu [42] attached the magnet to the animals’ scalp and secured it with a plastic band.

The main applications of the magnetic targeting process identified by the studies in question were magnetically targeted drug delivery [40], [41], [43], [47], [48], [49], [51], [52], [55], [56], [57], [58], [59], a new platform for monitoring magnetic resonance and targeted drug delivery [42], increased therapeutic efficacy of chemotherapeutic agents [44], development of a nanobubble system for cancer treatment [45], early cancer detection and more targeted treatment [46], more effective and tolerable treatment for cancer [50], applicability of inductively coupled plasma-optical emission spectrometer (ICP-OES) and electron spin resonance (ESR) studies in biodistribution studies of magnetic nanoparticles [53], and effective targeted drug delivery to the brain [54].

Figure 3 shows a schematic of the magnetic targeting process for tumor induction through different routes of SPION administration, the range of the dose used, and its delivery vehicle. In the same figure, the range and time of the external static magnetic field application for further evaluation by different techniques.

Figure 3:

Experimental designs of magnetic targeting process used in the studies selected for the review. *Number of repetitions of ExMF. Studies not mentioned in this figure did not present any of the data items outlined.

3.5 Toxicity

Although physicochemical characterization of SPION was described by studies included in our review of stability of SPION in vitro, the viability of the technique to apply SPIONs to tumor was discussed in a few manuscripts as well as the kinetic of biodistribution, elimination of SPION, and in vivo toxicity. These constitute the fundamental elements of the magnetic targeting process [47], [51], [53].

External imposition of a static magnetic field for tumor targeting by SPION is better observed in studies on the distribution of SPION in the organism, highlighting the higher concentration of SPION located in the tumor region and its microenvironment. Such fact proved the efficiency of this technique compared with tumor targeting without external magnetic field. Distribution data suggest, indirectly, the viability of the procedure given that the results showed that the physiological route of iron metabolism is maintained, and SPIONs are pushed by the organism, thus, reducing the long-term accumulation of iron in other organs that would turn them toxic [55], [59], [63].

Some selected studies, in addition to magnetic targeting technique application, have also applied therapeutic techniques using colorimetric and fluorescent assays and have shown a reduction of in vivo or in vitro tumor activity. This reduction was observed after tests with different combinations of SPION with different doses of chemotherapy agents and by imposition of external magnetic fields by alternation (generating heat in the tumor area) or not. These studies suggest that specificity of drug delivery to the local of interest using magnetic targeting resource contributes with efficiency to the therapeutic process [40], [43], [44], [45], [47], [51].

Despite the clinical and scientific knowledge on the advantages of the low toxicity that therapies with SPIONs present, because of its biocompatibility and possibility of controlling its migration by magnetic resonance imaging, further detailed studies are needed on the biological markers of free radical formation and viability [54]. The imposition of external magnetic field increases the speed and concentration in the migration route of SPION by the bloodstream, and this would cause a toxic effect in the organism. For this reason, detailed analyses are required to guarantee procedure precision to the therapy [54].

In addition, there is also the need to evaluate if imposition of external magnetic field to tumor targeting with SPIONs, even if static, do not cause excessive increase in temperature of acceptable biological heat band, fluctuations of blood pressure, and other secondary effects, therefore, enabling the continuity of homeostasis of the organism.

3.6 Analysis methods

The analysis of the efficiency magnetic targeting in the tumor region can be performed by a set of techniques such as histological assays, immunohistochemical, in post mortem tissues, or ex vivo and monitoring in vivo by MRI. In short, the selected studies address various neuroimaging techniques, illustrated in Table 5, from in vivo and ex vivo techniques.

Most studies evaluated the magnetic targeting efficiency by MRI technique [37], [38], [39], [41], [42], [43], [44], [45], [47], [48], [49], [50], [51], [52], [53], [54], [56], [57], [58]. Of these, 11 studies [39], [41], [42], [43], [47], [48], [51], [53], [56], [57], [58] also used as an analytical method electron spin resonance (ESR) spectroscopy, transmission electron microscopy (TEM) [39], [43], [49], [50], [54], [58], energy-filtered transmission electron microscopy (EFTEM) [39], focused ultrasound (FUS) [37], [44], [50], [54], fluorometry [44], inductively coupled plasma-mass (ICP-Mass) [44], Evans blue (EB) [37], [44], [54], hematoxylin-eosin (HE) [37], [38], [44], [50], Prussian blue (PB) [37], [38], [49], [50], [54], nuclear fast red (NFR) [48], [54], immunohistochemistry [50], inductively coupled plasma optical emission spectroscopy (ICP-OES) [38], [50], [53], [54], high-performance liquid chromatography (HPLC) [38], confocal/fluorescence microscopy [37], [38], [39], [54] or 4′,6-diamidino-2-phenylindole (DAPI) [54], and terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) [38]. Only four studies did not use MRI with an analytical method [40], [46], [55], [59]. Of these studies, two [40], [46] used fluorescence tomography, and one [55] used intravital microscopy. Huang [55] reported the use of ICP-OES, HE, PB, and NFR, and another study [59] reported the use of MET. Alternative or complementary methods to MRI were mostly performed ex vivo [37], [38], [39], [40], [41], [42], [43], [44], [47], [48], [49], [50], [51], [53], [54], [55], [56], [57], [58], [59] using tissue samples for ESR spectroscopy, TEM, fluorometry, ICP-mass, Evans blue, HE, PB, NFR, immunohistochemistry, ICP-OES, and DAPI. Some studies also carried out in vivo analyses, such as fluorescence [40], FUS [37], [44], [50], [54], and confocal intravital microscopy [46]/fluorescence microscopy [54].

Among the selected studies, MRI was the main method used for in vivo monitoring and targeting by SPION mediated by an external magnetic field to the tumor region. In addition to the accurate images of the different parts of the body, this technique is also advantageous because it is minimally invasive. The intensity range of the MRI equipment used was from 3.0 T [38], [49], [50], [54] to 7.0 T [37], [39], [41], [42], [43], [44], [45], [47], [48], [51], [52], [53], [56], [57], [58]. The most common sequences used were fast-spin-echo (FSE) [41], [42], [43], [47], [48], [51], [52], [53], [56], [57] and gradient echo (GRE) [39], [41], [43], [44], [47], [48], [50], [51], [52], [53], [56], [57], [58]. Some studies acquired images only 1 h after magnetic targeting [38], [41], [42], [43], while others did temporal measures between 1 and 240 h after targeting [37], [39], [52], [57], [58].