Animal Tumor Models for Boron Neutron Capture Therapy Studies (Excluding Central Nervous System Solid Tumors)

Introduction

Boron neutron capture therapy (BNCT) is a selective particle radiotherapy that combines the administration of boron compounds incorporated preferentially by tumor cells, followed by irradiation with thermal or epithermal neutrons. The capture reaction between a ¹⁰ B nucleus and a thermal neutron produces high-energy α-particles and recoiling ⁷ Li nuclei. These particles have high linear energy transfer, high relative biological efficacy, and a short path length of approximately the diameter of one cell, inducing a direct damage in DNA, causing irreparable double-strand breaks.^1–3 Thus, damage is circumscribed to tumor cells, where boron is preferentially localized, sparing normal surrounding tissue. In this sense, as BNCT involves biochemical rather than geometrical targeting, it would be potentially useful to treat all tumor cell populations, including undetectable micrometastases and foci of malignant transformation in field-cancerized tissue.^1,4 BNCT clinical studies around the world have shown therapeutic efficacy, associated with an improvement in patient quality of life and prolonged survival.^5–9 However, there is still room for improvement.

An ideal boron carrier should be nontoxic at therapeutic dose levels, should accumulate preferentially in tumor cells versus blood and normal tissue and should target all tumor cell populations. Two boron-containing compounds are routinely used clinically: boronophenylalanine (BPA) and mercaptoundecahydrododecaborate (BSH, sodium borocaptate). BSH is a diffusive drug, first employed in Japan in BNCT studies for patients with glioblastoma multiforme.^10,11 BSH is incorporated in brain tumors passively due to the disruption of the blood–brain barrier (BBB) in tumor. ¹² In the case of BPA, it was first introduced in BNCT for the treatment of melanoma, also in Japan. ¹⁰ BPA is incorporated by tumor cells by the L-type amino acid transporter 1 (LAT1). LAT1 could contribute to tumor growth by increasing the supply of amino acids and was found highly expressed in various cancers. ¹³ Apart from melanoma, BPA/BNCT has been used clinically for the treatment of other pathologies, for example, malignant brain tumors, head and neck cancer, mesothelioma, vulvar melanoma and genital extramammary Paget's disease, and colorectal liver metastases.^{7,9,14–16,18}

Boric acid (BA) and GB-10 (sodium decahydrodecaborate) are both diffusive boron agents and have also been studied for many years as possible boron agents to be used in BNCT for humans. GB-10 was first shown to be nontoxic in dogs and proposed as a boron agent for BNCT for malignant brain tumors and BNCT-enhanced fast neutron therapy. ¹⁹ GB-10 was approved for its use in patients by the U.S. Food and Drug Administration, but has never been evaluated in clinical BNCT trials like BPA and BSH due to disappointing results. A retrospective study showed that these results were due to a poor understanding of the boron agent biodistribution requirements and treatment management. ¹⁰ Because GB-10 has no special handling or storage requirements and it is nontoxic, it has recovered its interest and was employed in several translational experiments.^19–22 GB-10 does not traverse the intact BBB but is incorporated preferentially to brain tumors surrounded by a faulty BBB. Although GB-10 is not incorporated selectively to tumors in tissues with no BBB (such as oral mucosa tumors), GB-10/BNCT does exert a selective effect on tumors through a selective action on aberrant tumor blood vessels. ²⁰

BNCT mediated by BPA, BSH, GB-10, and BA have been widely explored in several animal models for different pathologies. ²³ The international BNCT community is making huge efforts to develop new boron delivery agents, for example, monoclonal antibodies, liposomes, nanoparticles, boron cluster agents, carrier proteins conjugated with boron compounds and bimodal drugs (with antitumoral and BNCT effects), such as metallocarboranes, among many others.^24–28

Translational research in adequate experimental models is necessary to optimize BNCT for different pathologies, particularly for those that respond poorly to standard therapies and would potentially benefit from more effective and selective therapies. Multiple radiobiological in vivo studies have been performed in ectopic, orthotopic, and metastatic cancer models (with implanted syngeneic or human cells or tumor pieces) and in chemically induced animal models like the hamster cheek pouch oral cancer model. Multiple boron compounds, routes of compound administration, and a range of treatment strategies were assessed in these animal models.^23,29 They are essential for the study of BNCT in terms of therapeutic efficacy and radiotoxicity, and to support the design of novel, safe, and effective clinical BNCT protocols. Despite the inevitable limitations of in vivo models, they contribute to the knowledge of the different aspects that are present in a clinical scenario.

This article will describe some examples of these radiobiological studies performed in different animal models and pathologies, highlighting the importance of experimental animal models for the advancement of BNCT. This article will not include in vivo models of central nervous system tumors, as they are addressed in another article of this special issue.

Chemically Induced Cancer Model: The Hamster Cheek Pouch Oral Cancer Model

A significant amount of work has been done in the chemically induced oral cancer model in hamster. Cancerization of the hamster cheek pouch closely mimics events involved in the development of precancer and malignant human oral lesions. Different cancerization protocols have been assessed: topical application of DMBA (dimethyl-1,2-benzanthracene) in mineral oil (0.5%) twice a week, during 6 weeks/8 weeks/12 weeks. The 6-week cancerization protocol allows for the study of the inhibitory effect of BNCT on the development of new tumors in field-cancerized tissue. ²¹ The 8- and 12-week protocols induce a tumor model surrounded by precancerous tissue, which in turn gives rise to the formation of additional tumors. Thus, these models are amenable to study both therapeutic efficacy and radiotoxicity in terms of mucositis in dose-limiting precancerous tissue, as occurs in field-cancerized oral mucosa in head and neck cancer patients. BNCT mediated by BPA (BPA/BNCT) induced a significant tumor response. Although BPA/BNCT induced mucositis, it resolved by the end of the follow-up.^20,30–32

BNCT-induced mucositis affects patient quality of life and treatment. The hamster cheek pouch is a widely accepted model to study mucositis. BPA/BNCT has been studied in animals exposed to different cancerization protocols (6-, 8-, and 12-weeks). BPA/BNCT-induced mucositis increased with the aggressiveness of the cancerization protocol employed in the hamster cheek pouch. In this sense, the study of BNCT in different oral cancer patient scenarios would contribute to personalize BNCT for each patient condition. ³³ In the oral precancer model induced with the 6-week protocol, the role of radioprotectors to reduce or prevent mucositis in the precancerous tissue has also been studied. This would allow higher doses to be delivered to the tumor, protecting the surrounding dose-limiting tissue. ³⁴

Biodistribution studies could suggest if a boron compound could be worth exploring for BNCT, depending on its gross boron concentration in the tumor and tumor/normal tissue, tumor/blood boron concentration ratios. However, radiobiological studies are essential to demonstrate the therapeutic effect of BNCT mediated by a particular boron compound/strategy and to evaluate its toxicity. For example, although the boron compound GB-10 does not accumulate in hamster cheek pouch tumors selectively, it did induce a high overall selective tumor response, damaging aberrant tumor blood vessels selectively, while sparing precancerous and normal tissue. ²⁰

Boron localization studies in different structures of the tumor and surrounding tissue are of importance to evaluate the potential efficacy of a boron compound and strategy. Studies performed in the hamster cheek pouch oral cancer model demonstrated that BPA accumulates in tumor parenchyma versus GB-10 that accumulates in stroma. ³⁵ Different strategies have been studied in the hamster cheek pouch to improve boron uptake and microdistribution in the tumor and enhance BNCT therapeutic effect. For example, Garabalino et al ²² demonstrated that the combination of electroporation with GB-10 increases boron uptake and optimizes GB-10 microdistribution in the tumor.

Another strategy such as the double application of BNCT (full-dose retreatment) mediated by the combination of GB-10+BPA inhibited significantly the development of second primary tumors from precancerous tissue upto 8 months after BNCT in the hamster oral precancer model (6-week cancerization protocol). In terms of radiotoxicity, the second application of BNCT did not exacerbate the mucositis observed after the first treatment in precancerous tissue.^36,37 Sequential BNCT (BPA/BNCT followed by GB-10/BNCT with an interval of 24 or 48 h) significantly enhanced tumor response versus a dose-matched single application of BNCT mediated by BPA+GB-10 in the classical oral cancer model (12-week cancerization protocol). ³⁸

The hamster cheek pouch oral cancer model is also useful to study transient normalization of defective blood vessels using thalidomide. Tumor blood vessel normalization using antiangiogenic agents would make tumor blood vessels less leaky, less dilated, and less tortuous, improving boron targeting and BNCT efficacy. ³⁹ The combination of thalidomide and administration of BPA in the normalization window improved boron targeting homogeneity in tumors, enhancing tumor response. ³⁹ When both strategies were combined (Seq-BNCT+Thalidomide), a 100% tumor response with 87% complete tumor remission was reached, with no normal tissue toxicity and no cases of severe mucositis in dose-limiting precancerous tissue. ⁴⁰

Ectopic, Orthotopic, and Metastatic Tumor Models

Apart from the chemically induced oral cancer model in the hamster, as the authors mentioned before, BNCT was studied in tumors that developed following implantation of tumor cells or tumor pieces in immunocompetent and immunodeficient animals. Ectopic tumor models are generally generated by injecting cells subcutaneously into the hind leg or back of the animals. In this sense, the injection site is different from the origin of the cultured cells. In the case of the orthotopic tumor model, cells are injected into the same origin site of the tumor. In the metastatic cancer model, cells are injected in vessels and lymph nodes causing metastasis at sites that are amenable to invasion. ⁴¹

In squamous cell carcinoma models in mice, plenty of examples can be found in the literature of different strategies that enhance boron uptake, microdistribution, and BNCT therapeutic effect on tumors, that is, the combined effect of BPA and BSH ⁴² ; the administration of a selective inhibitor of tumor blood flow, flavone acetic acid (FAA), with BSH: FAA would cause vascular collapse entrapping BSH in the tumor, inhibiting the clearance of BSH from the tumor with little or no effect in normal tissue ⁴³ ; and combination of the vascular targeting agent ZD6126 with BNCT, which enhanced the sensitivity of the quiescent (Q) tumor cells compared with the total tumor cells. Q cells are viable and clonogenic, but cease cell division in part because of hypoxia and depletion of nutrition in the tumor core. Therapies kill more efficiently proliferative rather than Q cells, the latter being responsible for recurrences after treatment. ⁴⁴ Sonoporation combined with BPA/BNCT ⁴⁵ and the combination of low-dose gamma irradiation with BNCT increased BPA accumulation in tumor and the tumor/normal tissue and tumor/blood boron concentration ratios, enhancing BNCT efficacy and extending the overall survival rate. ⁴⁶

In an ectopic poorly differentiated thyroid carcinoma model in mice, BPA/BNCT was proved therapeutically useful.^47,48 Moreover, the administration of a radiosensitizer like sodium butyrate increased tumor boron concentration after BPA administration and enhanced BPA/BNCT therapeutic effect, resulting in 80% of the animals with complete remission. ⁴⁹

BNCT was studied in two ectopic bladder tumor in vivo models (murine and human muscle invasive and high-grade bladder cancer cell line). A carbohydrate mimetic peptide that binds to the Anxa1 N-terminus (a target for the tumor vasculature) conjugated with BPA and BSH was studied. Yoneyama et al ⁵⁰ showed a lack of toxicity in these animals and a significant increase in BNCT therapeutic effect and boron compound stability. Yoneyama et al ⁵⁰ also demonstrated that BNCT induced tissue necrosis in bladder tumors accompanied by infiltration of CD8α-positive lymphocytes. This result suggests that BNCT induces an immune response by the host against tumor cells, triggering a strong cytotoxic reaction. BNCT-induced immune response was also shown in an ectopic breast cancer model in mice. BNCT was capable of switching peripheral blood mononuclear cells to an antitumor phenotype, together with a long-term effect on tumor growth when boron-rich liposomes were administered systemically. ⁵¹

In an ectopic colon cancer model in BDIX rats, the abscopal effect of BNCT was demonstrated for the first time. The abscopal effect refers to the inhibitory action of radiotherapy on the development and growth of nontargeted tumors, that is, at a site distant from the area of irradiation. In this model, the animals are first inoculated with syngeneic colon cancer cells in the right hind flank. The tumor is treated with BPA/BNCT and, 2 weeks post-BNCT, the animals are inoculated with these cancer cells in the left hind flank. BNCT reduced the tumor on the right leg and induced an abscopal effect on the left leg, as the authors observed that the nonirradiated left tumor was significantly smaller than the left control tumor. ⁵² Recent studies demonstrated that this immune response could be enhanced by the combination of BNCT with Bacillus Calmette–Guerin (BCG), an immune stimulator. BNCT+BCG induced a powerful abscopal response, a local effect and regional effects in tumor-draining lymph nodes. ⁵³

Besides, the ectopic colon cancer model in rats is also useful to study BNCT-induced dermatitis.^52,53 In this model, Trivillin et al ⁵³ reported a radiation dermatitis scale that was proved useful to study, after BNCT, the peak of dermatitis and healing time in the skin, the dose-limiting tissue in this study. Finally, in an autoimmune disease animal model of rheumatoid arthritis in rabbits, boron neutron capture mediated by BPA and GB-10 (boron neutron capture synovectomy [BNCS]) was also useful to treat the pathological synovium knee joints. The boron compounds were injected intraarticularly. BPA/BNCS and GB-10/BNCS successfully treated the pathological synovium in a model of antigen-induced arthritis in rabbits. ⁵⁴

BNCT was studied for the treatment of lymphosarcoma in a syngeneic ectopic model in Wistar rats. Masutani et al ⁵⁵ showed that a proinflammatory ligand called HMGB1 (high mobility group box 1 proteins), involved in the DNA damage response and cell death, and poly(ADP-ribose) a marker for both single- and double-strand breaks, were proposed as early and late markers of BNCT response. Particularly, in a more recent study in a xenograft mouse oral cancer model, the authors demonstrated that HMGB1 leaked out from tumor tissue as they could measure high levels of plasma HMGB1 on day 3 post-BNCT irradiation. This study shows that tumor response to BNCT could be monitored by measuring DNA damage biomarkers in plasma of patients. ⁵⁶ Biomarkers for early and late responses to BNCT will be useful for optimizing the conditions of BNCT, biological evaluation of new boron delivery agents, and to predict the therapeutic efficacy and toxicity of BNCT alone or combined with other strategies.

Several studies have been performed to optimize BNCT for melanoma. BPA was first studied as its chemical structure is similar to that of tyrosine, which is required for melanogenesis. Coderre et al^57,58 demonstrated in mice that BPA is taken up by melanoma tissue to a much greater extent than by normal tissues. Mishima et al ⁵⁹ showed a significant therapeutic effect of BPA/BNCT in melanoma-bearing hamsters. The results led to study of BPA/BNCT in spontaneously occurring malignant melanoma in Duroc pig skin. Mishima et al ⁵⁹ cured melanoma with minimal damage to the surrounding skin. Morita et al ⁶⁰ showed that intratumoral injection of the tyrosinase gene increased BPA accumulation in the tumors, inducing an enhancement of the tumor-suppressive effect of BNCT on amelanotic melanoma in hamsters.

BNCT for the treatment of colon cancer has also been explored in ectopic syngeneic colon cancer models in mice. For example, Kikuchi et al ⁶¹ compared BSH versus MID-BSA (maleimide-functionalized closo-dodecaborate [MID] conjugated to bovine serum albumin [BSA]). MID-BSA accumulates principally in tumor tissue due to the enhanced permeability and retention effect. In this study, they observed high therapeutic effect in tumors by MID-BSA/BNCT and although a high level of boron was circulating in blood during thermal neutron irradiation, no serious damage was observed in the irradiated area after irradiation.

In the same animal model, Tsurubuchi et al ⁶² studied the biodistribution of a boron-containing α-d-mannopyranoside by in vivo fluorescence imaging and reported a highly significant therapeutic effect on tumors post-BNCT. This boron compound accumulates in tumor due to the Warburg effect (upregulated anaerobic glycolysis). Finally, it is well known that colon carcinoma induces liver metastases. In a syngeneic colon carcinoma liver metastases model in BDIX rats (induced by a subcapsular inoculation of syngeneic colon cancer cells in the liver), BPA/BNCT induced a significant therapeutic effect on tumors, with no liver radiotoxicity.^63,64

In vivo hepatic cancer models have also been used to study the therapeutic effect of BNCT. For example, Lin et al ⁶⁵ demonstrated that BNCT mediated by BA induced tumor control in rats, with no damage to normal liver tissue. In a syngeneic hepatic cancer model in rabbits (rabbit tumor cells were injected into the left lobe of their liver), BNCT mediated by BSH entrapped in TF–PEG liposomes (Transferrin [TF] conjugated polyethylene glycol [PEG] liposomes) with ¹⁰ B-distearoyl-boron lipid demonstrated a significant inhibitory effect on tumor, with nuclear deformation and apoptotic bodies after BNCT. ⁶⁶

In rabbits with multifocal liver tumors, Hung et al ⁶⁷ demonstrated the safety of BNCT mediated by BA in two fractions (2 days between fractions). They observed no obvious damage to the hepatocytes or vessels in the normal liver regions, while tumor cells and vessels in tumor masses were damaged and underwent fibrosis as well as necrosis. Yang et al ⁶⁸ implanted tumor pieces in the liver, instead of injecting tumor cells. They assessed BA microdistribution in the tumors and BNCT therapeutic effect and observed that BA accumulated preferentially in blood vessels and that BA/BNCT induced tumor vascular injury.

The possibility to use BNCT to treat lung cancer and metastases in the lung was extensively studied in a wide variety of animal models. For example, Farías et al ⁶⁹ reported biodistribution studies using BPA in an adult sheep model and performed computational dosimetry studies for a human explanted lung to evaluate the feasibility and the therapeutic potential of ex situ BNCT. Boron microdistribution studies using BPA were also performed in the lung metastases model in BDIX rats. ⁷⁰

Masunaga et al ⁷¹ studied the effect of bevacizumab on local tumor response and lung metastatic potential during BNCT in melanoma tumor-bearing mice. They observed that bevacizumab administration (an antivascular endothelial growth factor antibody, that inhibits tumor vascular growth) and the combination with mild temperature hyperthermia (MTH), but not with nicotinamide (an acute hypoxia-releasing agent), further enhanced total tumor cell population sensitivity. Masunaga et al, ⁷² in the same animal model, demonstrated that BSH/BNCT combined with a hypoxic cytotoxin tirapazamine (TPZ), with or without MTH, improved local tumor control, while BPA/BNCT in combination with both TPZ and MTH as well as nicotinamide reduced the number of lung metastases.

Bakeine et al ⁷³ described another interesting model, in which a suspension of syngeneic colon carcinoma cells preirradiated with BPA/BNCT were inoculated in the inferior vena cava of BDIX rats. In vitro BNCT-treated tumor cells have a significantly reduced capacity to induce lung metastases in rats (in this case, the lung was not irradiated). In a syngeneic colon carcinoma lung metastases model (cells injected in the jugular vein of BDIX rats), a significant reduction in the number of metastases was observed after BPA/BNCT with no significant radiotoxicity in the lung. ⁷⁴

Clear cell sarcoma (CCS) was also studied as a possible target for BNCT. Human CCS cells were injected in nude mice, in two different sites: dorsal or femoral. Biodistribution studies using BPA showed that dorsally tumor-bearing mice had significantly higher peak boron concentrations in tumor, compared with the femorally tumor-bearing mice, due to differences in the extent of angiogenesis. ⁷⁵ In the same animal model, BPA/BNCT was therapeutically useful to treat these tumors, with no normal tissue radiotoxicity. Andoh et al^76,77 implanted the human CCS cell line in the parenchyma of the left lung in nude mice to model CCS lung metastases. Using this model, they performed biodistribution and autoradiography studies. The authors evaluated a two-fractionated neutron irradiation. As neutrons have poor penetration two portals were used to optimize dose distribution. BNCT made it possible to decrease the tumor volume in CCS lung metastases without inducing significant side effects. These findings imply that BNCT with whole lung neutron irradiation may be effective even for invasive tumors.

To study the effect of BNCT in bone metastases, a human breast cancer cell line was injected into the tibia of nude mice. After 8 weeks, the tumor developed in the animal's bone. In this animal model, Andoh et al ⁷⁸ observed differences in tumor response to BNCT: animals with pathological fracture did not respond to BNCT, probably due to heterogeneity in the distribution of boron in bone tissue. In the nonpathological fracture group, BNCT had antitumor efficacy specific to the bone tumor site, without adversely affecting the surrounding tissue. In addition, BNCT might promote bone regeneration. ⁷⁸

BNCT has proven to induce some therapeutic benefit in a mouse model of mesothelioma. A dual boron agent, sulfonamido-functionalized-carborane, which acts as carbonic anhydrase IX (CAIX) inhibitor and boron delivery agent, was tested in a mesothelioma animal model (mouse AB22 cells were injected subcutaneously in the neck of adult mice), showing a significant reduction in the tumor volume. CAIX enzyme is overexpressed in mesothelioma, as it contributes to cell survival in a hypoxic microenvironment. ⁷⁹ Other agents like hyaluronic acid containing BSH and theranostic agents targeting low-density lipoprotein receptors were also reported, showing an increase in the survival time of the animals and a significant reduction in tumor volume with no toxicity.^80,81

Pancreatic cancer was also evaluated as a potential target for BNCT. A human pancreatic carcinoma cell line, which expresses carcinoembryonic antigen (CEA), was injected subcutaneously into the back of male nude mice. Boronated immunoliposomes anti-CEA (anti-human CEA monoclonal antibody was conjugated to the liposome surface) were injected intratumorally and after neutron irradiation, a significant reduction in tumor volume was achieved. ⁸² Also for gastric cancer, boronated immunoliposomes were targeted with a monoclonal antibody (MGb2) and proved to be therapeutically effective against tumors. ⁸³ Gold nanoparticles labeled with ¹²³I conjugated with human epidermal receptor-2 antibody were used as theranostic agents in mice implanted with a human gastric cell line. Tumor accumulation in microSPECT images and biodistribution studies would reflect the boron content of the tumor, suggesting that this imaging platform would predict the treatment outcome. ⁸⁴ Animal models are also essential to elucidate if a boron compound has theranostic characteristics.

The first preclinical proof of principle that BPA/BNCT reduces the in vivo growth of a human prostate cancer cell line injected into the back of nude mice was reported by Takahara et al. ⁸⁵ Biodistribution studies using BPA were performed and they showed that BPA/BNCT induced a significant reduction of the tumor without any severe side effects. Other studies reported new boron agents that warrant analysis in this type of cancer, like different nucleoside-related derivatives ⁸⁶ and boranophosphate analogs of RNA for enhancing the potential of gene-specific anticancer strategies. ⁸⁷ Prostate-specific membrane antigen (PSMA) is a target for prostate cancer imaging and drug delivery. Several studies employing PSMA-targeted boron compounds and nanoparticles were tested in vivo.^88–90

Discussion and Conclusions

This review describes how translational studies in animal models optimized BNCT for different pathologies and their importance in the evaluation of new boron compounds and administration and treatment strategies. Animal models are useful for the study of the stability and potential toxicity of new boron compounds or delivery systems and show how BNCT theranostic strategies and the evaluation of biomarkers, before and after BNCT, can anticipate and monitor therapeutic and adverse effects. Another very interesting finding that can only be assessed in in vivo cancer models is that BNCT induces an immune response by the host against tumor cells. This immune response could be modulated to enhance the therapeutic effect of BNCT.

There are plenty of examples demonstrating that BNCT could be broadened to several pathologies that have not been studied in the clinical field yet. However, several of these studies have been mainly performed in ectopic cancer models. In this sense, there is a need for more BNCT radiobiological studies in chemically induced and/or orthotopic in vivo models to assess treatment planning and potential radiotoxicity in dose-limiting organs and tissues. For example, in the case of BNCT for pancreatic cancer, this illness is particularly challenging as the organ is located deep in the abdomen and surrounded by many sensitive organs. A robust knowledge of boron uptake and therapeutic effect of BNCT in tumors located in the pancreas would help to assess a treatment plan able to ensure the therapeutic action and to remain safe for the rest of the irradiated tissues.

Another example was mentioned in El-Zaria et al ⁸⁸ who commented on the problems that have to be solved before starting with clinical trials in prostate cancer. They suggested how to irradiate a prostate tumor, considering that neutrons can only reach organs close to the body surface and their effect decreases with increasing depth. In this sense, orthotopic tumor models would allow for the study of the safety and effectiveness of the suggested transperineal treatment plan. ⁸⁸ Other examples further support this notion. The study in the hamster cheek pouch oral cancer model showed that BPA/BNCT-induced mucositis in precancerous tissue (the dose-limiting tissue that surrounds the tumors) increased with the aggressiveness of the cancerization protocol. ³³ Furthermore, Andoh et al ⁷⁸ showed that a pathological fracture could affect the therapeutic effect of BNCT on breast cancer metastases in bone.

Animal research allows for the study of BNCT for different pathologies, both in terms of therapeutic efficacy and radiotoxicity. In vivo studies address the issues of tumor localization and tissue characteristics, including dose-limiting tissues around the target tissue, the role of vasculature, and the immune system. Also, many of the strategies employed to optimize BNCT are approved for their use in humans. These studies are of significant importance as they pave the way to design, in the near future, a more personalized BNCT for each patient and design safe and effective treatment protocols.

Much work has been done principally on head and neck tumors and glioblastoma, but also several studies showed that BNCT could be broadened to other pathologies. However, some of them, like pancreatic cancer, are targets that should be addressed carefully as dose-limiting organs (like Intestine) could limit the dose to the tumor. Also, deep seated tumors represent a limitation for BNCT, due to the poor penetration of the neutron beam in many cases. Nowadays, with the introduction of accelerators designed especially for BNCT and the study of new boron compounds and strategies based on tumor characteristics and location, these limitations would be easier to overcome. BNCT has a unique advantage: only one application, or at most two, demonstrated a significant increment in patient survival and quality of life. In some cases, BNCT cured the tumor, in a patient with no other possible treatment. However, it is not a systemic therapy. In this sense, future studies should be addressed combining BNCT with systemic therapies, aiming at the total recovery of the cancer patient.

All these radiobiological findings reported in the review suggest that BNCT studies in animal models would be essential to design novel, safe, and effective clinical BNCT protocols for existing or new targets for BNCT.