Individual Radiation Sensitivity and Biomarkers: Molecular Radiation Biology

Abstract

In recent years, scientific understanding of the changes radiation makes to the various tissues of the body has vastly increased. Identification of biological markers of radiation exposure and response has become a wide field with an increasing interest across the radiation research community. This chapter introduces the concepts of individual radiosensitivity, radiosusceptibility, and radiodegeneration, which are the key factors to classify radiation responses. Biomarkers are then introduced, and their key characteristics as well as classification are explained, with a particular focus on those biomarkers which have been identified for use in epidemiological studies of radiation risk—as this is a crucial topic of current interest within radiation protection. Brief information on collection of samples is followed by a detailed presentation of predictive assays in use in different settings including clinical applications with responses assessed chiefly in tissue biopsy or blood samples. The sections toward the end of this chapter then discuss the evidence associated with the relationship between age and separately sex, and radiosensitivity, as well as some genetic syndromes associated with radiosensitivity. The final section of this chapter provides a brief summary of how our current knowledge can further support individual, personalized, uses of radiation, particularly in clinical settings.

FormalPara Learning Objectives

• To understand the different responses of tumor and normal tissues to ionizing radiation (IR) whether at low or high dose.

• To grasp the importance of radiation biomarkers of exposure as well as effect and their integration with molecular epidemiological studies.

• To be able to discuss current and emerging biomarker methods to predict normal tissue and tumor responses to radiotherapy (RT).

• To understand how age and sex influence IR sensitivity on cellular and individual levels as well as health risks induced by IR exposure.

• To grasp how genetic syndromes can be associated with an increased radiation sensitivity and cancer risk.

• To understand the concept of precision medicine in context of RT and future research avenues in this field.

7.1 Definition of Individual Radiosensitivity, Radiosusceptibility, and Radiodegeneration and Radioresistence

The term “radiosensitivity” is one of the most extensively used words in radiobiology. It was described as radiation-induced tissue reactions (e.g., skin is radiosensitive) in the first decade of the nineteenth century [1]. Since 1930s, with the first Congresses of Radiology, the term “radiosensitivity” was also used as a synonym of radiation-induced cancers (e.g., thyroid is a radiosensitive organ) and progressively was used for radiation-induced cataracts (e.g., eyes are radiosensitive [2]). All these different uses lead to an actual confusion and notably raise legal issues since radiation-induced cancers, cataracts, or skin burns do not correspond to the same level of clinical injuries [3].

To avoid these confusions, a possible approach is to consider all the major clinical features of the response to radiation by using unequivocal terms that could be indifferently applied to the individual, tissue, cellular, or molecular scales. To consolidate this approach, it is important to document the individual response to radiation through a complete knowledge of its different features. For example, it is noteworthy that ataxia telangiectasia (AT), caused by homozygous mutations of the AT mutated (ATM) gene resulting in aberrant ATM protein, is associated with post-RTfatal reactions and high risk of leukemia [4], while Li Fraumeni’s syndrome is associated with cancer proneness but not with significant post-RT adverse tissue reactions [5]. Conversely, Cockayne’s syndrome is associated with significant tissue radiosensitivity but no cancer proneness [6].

Hence, the following definition has been therefore proposed in literature [7] and summarized in Table 7.1.

Table 7.1 Summary of the terms—radiosensitivity, radiosusceptibility, and radioresistance

• “Radiosensitivity” is the proneness to radiation-induced adverse tissue events that are considered as non-cancer effects attributable to cell death. Radiosensitivity is generally correlated with unrepaired DNA damage and observed in response to high doses of radiation [8] (Box 7.1).

• “Radiosusceptibility” is the proneness to radiation-induced cancers which are non-toxic effects attributable to cell transformation and/or genomic instability (in part correlated with DNA misrepair). Since IR is considered to be a carcinogenic agent, radiosusceptibility is distinctly and strongly linked to susceptibility to spontaneous cancer induction. The term “radiosusceptibility” was proposed due to its similarities with “cancer susceptibility,” extensively used in the ICRP (International Commission on Radiological Protection) reports and since it introduces the notions of stochastic events [9].

• “Radiodegeneration” responses are non-cancer effects attributable to mechanisms which are related to accelerated aging and often correlated with unrepaired DNA damage that is tolerated by and accumulated in cells [9]. Radiodegeneration responses cannot be considered like radiosensitivity responses as defined above since their incidence rates, the types of cellular death, and the genes involved are different.

Box 7.1 Cellular Factors Influencing Radiosensitivity

• Radiosensitivity differs throughout the cell cycle with, in general, G₁ phase taking an intermediate position, and late S phase being most radioresistant.

• The greater proportion of repair by HR than by NHEJ in late S phase may explain the resistance of late S phase cells.

• Chromatin compaction and poor repair competence (reduced enzyme access) could explain the high radiosensitivity in G₂/M.

7.2 Biomarkers of Radiation: General Considerations

7.2.1 Definition

A biomarker is an objective feature with one or more defined characteristics which indicate specific normal biological and pathological processes, or responses to an exposure or to therapeutic interventions. To date, radiation biomarkers are primarily identified in blood or saliva and are measurable indicators that reflect an interaction between a biological system and one or more environmental agents (chemical, physical, or biological). Biomarkers provide crucial information on the complex molecular cascade of events and their mechanisms underlying the pathological conditions or pharmacological responses to a therapeutic intervention.

Biomarkers can be used to assess various different types of biological characteristics or parameters. These include genetic sequences, receptor expression patterns, radiographic or other imaging-based measurements, blood composition, electrocardiographic parameters, or organ function. Since biomarkers are quantifiable, they can be used to characterize the response to direct or indirect IR exposure, to select radiation dose, and to assess the potential safety issues related to dose administration. A large number of such biomarkers have been developed over the years; the characteristics of the different classes of radiation biomarkers will be reviewed in Sect. 7.2.2.

7.2.2 Characteristics of a Good Biomarker

Although the definitions, nature, and use of biomarkers are multiple and rapidly evolving with the sophisticated—omics technologies, they must be evaluated in terms of their ability to address etiology and genetic susceptibility, predict and quantify dose of exposure. There are certain properties which are desirable when linking a biomarker with an exposure, e.g., IR. These include high specificity and sensitivity, known variability in the general population, should give reproducible results when assessed and multiplexing of analyses to allow for screening purposes. Some additional desirable characteristics of an ideal biomarker can be listed for use in large scale molecular epidemiological studies: (a) Early expressivity; (b) Linear relationship across time; (c) Strong correlation with a health effect; (d) Reproducible between laboratories; (e) Biologically plausible; (f) Inexpensive and feasible for sample collection; (g) Consistency (the same exposure will produce the same concentration of the biomarker every time).

7.2.3 Radiation Biomarkers for Potential Use in Epidemiological Studies

Radiation biology research has identified several approaches, especially the “omics” fields, as promising avenues for the development of suitable biomarkers of high sensitivity and specificity for radiation exposure. Radiation epidemiology biomarkers should preferably be specific to radiation and independent of other environmental exposures such as tobacco or cigarette smoke. Such a biomarker would simplify analysis and help to substantiate radiation causality. Though biological biomarker often lack specificity, they can still be informative in predicting the development of radiation-induced disease if such exposures are additive or interactive. Multi-biomarker approaches should be particularly useful in epidemiological studies, both for (1) assessing exposure–response relationships and how they vary with individual susceptibility and (2) to understand better disease mechanisms and the interplay of different possible pathways.

Contrariwise, carefully planned molecular epidemiological studies are crucial for the validation and verification of biomarkers, to determine their specificity and sensitivity as well as factors that might influence them (e.g., age, sex, smoking status, environmental agents, chronic conditions such as inflammation or individual sensitivity) [10].

7.2.4 Integrating Biomarkers into Molecular Epidemiological and Biological Studies

The ultimate goal of using biomarkers in molecular epidemiological studies is to be able to predict health risk. The types of biomarkers mentioned in Table 7.2 hold substantial prospective in epidemiological studies; however, there are a number of key questions to be considered which are generic to their application. Among these are: “What to measure?” “Where to measure?” and “When to measure?”

Table 7.2 Classification of biomarkers based on temporal parameters

Determining if a biomarker is a good biomarker (the characteristics to determine an ideal biomarker are outlined in Fig. 7.1) for molecular epidemiological studies is complex because this relies on a number of different concepts associated with the radiation exposure and otherwise, which will very much depend on the biological samples such as cells (buccal cells, fibroblasts, hair follicle cells, etc.), blood, saliva, urine, tooth, hair, nail which can be collected non-invasively at different time points post IR exposure to study biological effects.

Fig. 7.1

Characteristics to determine an ideal biomarker. An ideal biomarker for molecular epidemiological studies (top) and general considerations of a good biomarker (bottom)

There is a great interest in developing new biomarkers for radiation exposure, which could be used in large molecular epidemiological studies in order to correlate estimated doses received by individuals and health effects using high-throughput technologies, i.e., “omics.” In these instances, biomarkers can provide a measure of external exposure as well as internal absorption of radioactive material and can thus be markers of internal dose as well. However, factors other than the exposure may influence biomarker expression, and thus there may not always be a simple relationship between external exposure and internal dose. For example, DNA or protein adducts may be applicable as markers of other processes such as absorption, distribution, metabolism, and DNA repair, as well as of exogenous exposure. As a result, the measured internal radiation doses (biomarkers) will be an amalgam of exposure and these variables.

In a nutshell, reliable radiation biomarkers databases can be shaped by integrating the information from radiation genomics, metabolomics, and proteomic analysis in order to expand the scientific frontiers on predicting and/or monitoring radiation exposure-associated effects. Such protocols along with more sophisticated technologies are probably vital for the development of personalized medicine and will undoubtedly prove highly useful to bring a new horizon of therapeutic possibilities [12] (Box 7.2).

Box 7.2 Biomarkers for Epidemiology and Dosimetry

• Biomarkers and/or biological dosimeters are essential for predicting and/or monitoring radiation exposure-associated effects, quantifying the exposure, estimating absorbed radiation dose in certain accidental situations or a suspected radiation overexposure.

• Radiation epidemiology biomarkers should be specific to radiation and independent of other factors that might influence them (such as age, chronic conditions, smoking, tobacco, or individual sensitivity).

• Identifying biomarkers of IR exposure employs a multi-parametric approach to achieve an accurate dose and risk estimation.

7.2.5 Biological Classification

Several biological responses can act as potential biomarkers for IR exposure. They are linked to cellular or physiological mechanisms which have been shown to change soon after radiation exposure. The use of various “omic” technologies together with hypothesis-driven approaches may be highly useful to measure radiation biomarkers in a biological system. Some biomarkers could be used as response markers or as surrogate endpoints to predict radiation side effects. The expression levels of many biomarkers can be expected to be correlated with each other and so could be classified in multiple categories, such as

• Phosphorylated histone H2AX (γ-H2AX) acts as protein biomarker for radiation exposure but is useful as a DNA damage marker; suggesting a close one-to-one relationship between initial as well as residual radiation-induced DNA DSBs and γ-H2AX foci.

• 8-oxo-dG acts as a marker of nucleotide damage but is strongly associated as a maker of oxidative DNA damage suggesting it is produced abundantly in DNA exposed to free radicals and reactive oxygen species (ROS).

• Phosphoproteomic profiling insights into processes influenced by epigenetic modifications, but also uncovers signaling pathways.

These biomarkers can be organized in categories such as (a) cytogenetic; (b) nucleotide pool damage and DNA damage; (c) germline inherited mutations or variants; (d) induced mutations; (e) transcriptional and translational changes; (f) epigenetic modifications; (g) lipid peroxidation; (h) others, including biophysical markers (Fig. 7.2).

Fig. 7.2

Biological classification of radiation biomarkers. (Reproduced with permission, with some modification (changed layout and some content), from [10]; licensed under CC BY-NC-ND 3.0)

Biomarkers of radiation exposure are further discussed in Chaps. 2, 3, and 8.

7.3 Temporal Classification of Biomarkers

Over the past decades, the definition and classification of the different types of biomarkers have varied slightly, depending on the biomedical field considered. Biomarkers are an important aspect of radiation countermeasure development and can be used as a trigger for intervention as well as in selecting a radiation dose and treatment regimen in humans, e.g., in the context of RT of cancer ([13] and as further discussed here in Chap. 7 and in Chap. 8). Biomarkers can also provide information on potential modifying/confounding factors to allow an assessment of biological interactions. Pernot et al. [10], Hall et al. [14] classified biomarkers into four broad categories, based on their temporal parameters Table 7.2.

One should be cognizant of the fact that overlap does exist between these different types of biomarkers. Taken together, these attributes enable a better understanding of exposure and its effect on biological pathways across different forms of exposure, health changes, disease headway; providing more meaningful comprehensive risk assessment (Fig. 7.3). This classification is acceptable not only with respect to the timing of processes that can be measured with these biomarkers, but also in considering the most adequate designs and sampling procedures in molecular epidemiological studies (Box 7.3).

Fig. 7.3

Timeline of radiation-induced disease progressions and relation with different types of radiation biomarkers. (Reproduced with permission, with some modification (changed color and layout), from [10]; licensed under CC BY-NC-ND 3.0)

Box 7.3 Use of Biomarkers at Different Times Post Exposure

• Biomarkers of exposure are available at some point after exposure and are suitable for estimating the dose received.

• Biomarkers of susceptibility can be available before, during, or after exposure and can predict an increased risk of radiation effects.

• Biomarkers of late effects can be used to assess health effects that are present a long time after exposure before clinical detection of the radiation induced disease or death.

• Biomarkers of persistent effects allow the assessment of radiation effects present a long period of time after exposure.

7.4 Collection of Individual Samples for Radiation Studies

Not only since the “age of OMICS,” it is well known that the collection of samples from patients and healthy volunteers is essential for future research [15]. Especially in radiation research, the collection of biological samples represents a significant part of translational research, since, especially in studies on radiation protection, the collected biological samples represent an essential parameter for analysis of (historic) radiation exposure [14]. Also, in therapeutic trials, the sampling of tissues and body fluids has an immense impact on the search, discovery, and validation of novel biomarkers supporting the pathological diagnosis as well as for the determination of therapy toxicity and/or outcome [15, 16]. The collected samples may safely expand the possibilities of the entire analytical process within prospective and retrospective trials in radiation science. However, the collection and storage of biological samples should always be done within a quality-controlled manner [17]. In contrast to tissue sampling, the sampling of body fluids also has the big advantage, that these samples are nearly always available or easy to access like vein puncture. In addition, the sampling of blood can be done mostly together with clinical mandatory blood draws, minimizing the burden of the patient/donor, and enhancing the amount of time points of sampling as well as it also increases the donor’s acceptance of giving blood for research. Since the collection of samples at biobanks accelerate the process of transferring scientific knowledge into therapeutic application, it should be the duty of clinical researchers adding translational programs with sample collection to the prospective clinical trials. In our hands, the sampling and analysis of immunological parameters lead to predictive biomarkers supporting the pathological diagnosis and therapeutic intervention and helping radiation treatment in precision medicine [16, 18].

The sampling, the processing, as well as the storage have to be done in a quality-controlled manner as outlined before by Winter and colleagues [19] or at the respective international biobank consortia like the European, Middle Eastern and African Society for Biopreservation and Biobanking (ESBB), or the European research infrastructure for biobanking (BBMRI-ERIC; [20]). There are another three very important things to keep in mind when collecting samples for later analysis: The best collections are almost worthless if they are not connected with clinical, radiation exposure, and patients/donor data. Along with this, the informed consent of the donor should allow use of the samples for the respective analyses even when the samples will be given for, e.g., “OMICS” analysis to a cooperation partner or to the statistician who performs the analysis of the data. Lastly, a biobank is a living “thing” which should be used for research and not as a secure vault to store samples for eternity. Taken together, it should be the duty of research on humans to collect and store samples for further analyses, but also processing and storage should be carried out in accordance with applicable regulations and standards. This is the only way to ensure that the samples do not suffer any loss of value and are available for later applications.

7.5 Predictive Assays

Tissue reactions induced by IR are the result of different types of cell death (mitotic death, apoptosis, autophagy, senescence, etc.). Loss of clonogenicity (and not physical disappearance or metabolic shutdown) appeared to be common to all types of cell death.

In 1957, Puck and Markus proposed to use clonogenic assays (or colony method) to quantify cellular radiosensitivity [21] (Fig. 7.4). The assay is based on the ability of an individual tumor cell to grow into a colony after exposure to various doses of radiation given the surviving fraction (SF) at each dose. The fraction of cells surviving after 2 Gy (SF2) has been demonstrated to be a robust predictor for radiation sensitivity but colony formation can take 7–14 days. Several predictive tests have been proposed to approach clonogenic survival but with a suboptimal statistical power.

Fig. 7.4

Brief overview of the original clonogenic assay

The original clonogenic assay can be performed in two different ways (1) irradiation after plating or (2) plating after irradiation. The first is usually carried out to investigate intrinsic radiosensitivity to varying modalities (types) of treatment, and the second allows for the assessment of reproductive ability. The irradiation after plating method is presented in Fig. 7.4 as it is usually used in radiobiological studies. Further details on the clonogenic assay can be seen in Chap. 3.

Several techniques continue to rely on cell culture:

• The level of radiation-induced micronuclei (MN) has been quantitatively correlated with radiosensitivity since the 1960s thanks to a simple and robust protocol consisting of blocking the process of cytokinesis by drugs such as cytochalasin B [22].

• The premature chromosome condensation (PCC) assay consists in making chromosome fragments appear more quickly by fusing the tested cell with a cell in mitosis. The heterokaryon thus formed allows the exchange of mitotic factors in the cell into G₀/G₁ and then produces premature condensation of the chromatin [23].

• The enumeration of chromosomal aberrations by fluorescence in situ hybridization technique.

It made sense to focus on the molecular mechanisms of the cellular response to radio-induced damage to DNA and in particular to DNA repair, since a quantitative link between unrepaired Double-Strand Breaks (DSB), chromosomal breaks, and radiation-induced clonogenic cell death was demonstrated [24]. Moreover, the vast majority of genetic syndromes associated with individual radiosensitivity are linked to mutations in genes involved in radiation-induced DSB signaling or repair. DSB measurement techniques were investigated with some confusion on their specificity and instead reflected other types of damage [7]. The first techniques for measuring DSB were based on discriminating radiation-induced DNA fragments based on their size. This was particularly the case with sedimentation in sucrose gradients, neutral elution, and pulsed-field electrophoresis. Such a principle has the advantage of measuring the repair of DSBs independently of any molecular repair pathways regardless of the post-irradiation time. On the other hand, these techniques do not make it possible to assess the quality of the repair, that is to say whether it is faithful or at fault.

The halo technique consists, using fluorescent intercalators, in quantifying such an increase in the nucleus. The comet technique combines electrophoresis and the halo technique, both applied individually to each cell. Data from the comet technique are usually given in the form of the product of the increase in the size of the nucleus (comet head) times the distance that DNA fragments migrate (comet tail) [25].

From 2003, with indirect immunofluorescence, it became possible to follow precisely and in real time in the nucleus and for a wide dose spectrum, the kinetics of appearance/disappearance of DNA repair proteins (foci). A correlation between SF2 and the rate of unrepaired DSB 24 h after 2 Gy (γ-H2AX marker) could be demonstrated—constituting a functional repair test [26]. A second marker significantly increased the performance of the test—based on the speed of nuclearization and the functionality of the pATM protein in the nucleus [8].

Genome-wide association (GWAS) studies have been widely used to identify associations between commonly occurring variations in DNA sequence, such as single nucleotide polymorphisms (SNP) and human traits such as individual radiosensitivity [27]. A few SNP as well as mitochondrial haplogroups have been reported.

The interest in the predictive power of the transcriptome began in the early 2000s and in vitro transcriptomic signatures for individual radiosensitivity then emerged. The identified predictive genes were associated with cellular functions such as TGFβ pathway, particularly extracellular matrix remodeling, apoptosis, proliferation, and ROS scavenging [28, 29] (Table 7.3).

Table 7.3 Reported gene expression signatures for individual radiosensitivity

Epigenetic modifications include histone modifications such as acetylations and methylations, DNA methylation, particularly on CpG island, non-coding RNAs, and three-dimensional chromatin organization. As this a relatively new field, only few studies have been conducted on the epigenetic regulations of skin fibrosis, mainly on miRNAs [32, 33].

Ozsahin et al. [34] developed a rapid radiosensitivity test (<24 h) based on lymphocyte apoptosis, a biological response developing 6–72 h after irradiation. The authors showed that low Radio-Induced Lymphocytic Apoptosis (RILA) was significantly correlated with late grade ≥2 tissue toxicities.

In recent years, the identification of routinely available blood and clinical markers that may help to predict the response to immune therapies alone and in multimodal settings including RT has been in focus of scientists of several disciplines and clinicians [35, 36]. Here, immune markers of the peripheral blood are key factors, since they circulate in the body and enter several tissues in response to disease, therapy, and stressors such as radiation. Stress and immune parameters should jointly be considered, and a differentiation between primary radiation signatures and consecutive systemic immune biosignatures is challenging, but anyhow interconnected [37, 38]. Notably, single immune parameters are insufficient, but rather immune profiles that reflect the complexity of the immune system and the manifold interactions of its cellular and soluble components [16, 39].

7.5.1 Predicting the Response of Tumors to Radiotherapy

Tumor response to RT is a multi-faceted metric of outcomes after radiotherapeutic treatment, often observed through biopsy of the tumor or liquid biopsy. There is currently no universal definition of tumor response, but it can generally be considered as any favorable response of the tumor to therapy. Tumor biopsies may aid in the development of personalized patient treatment regimens by providing molecular and structural material for use in developing metrics capable of identifying who will or who will not favorably respond to RT. For patients who do not respond favorably to treatment, they can be offered another more effective avenue of treatment, sparing them from treatment toxicity.

Identifying treatment-resistant phenotypes/genotypes and therapeutic targets that may influence tumor response is at the center of current radiation biology research. Robust, patient-specific, and predictive biomarkers are critical to assess tumor response to improve patient treatment and outcomes. There is an unmet clinical need to identify translational biomarkers that allow for tailor-made and optimized patient-specific treatment. Patient tumor samples such as tissue and liquid biopsies along with varying modalities of analysis that can be performed to identify predictive biomarkers of RT treatment response will be discussed in this subsection (Figs. 7.5 and 7.6).

Fig. 7.5

Patient tissue biopsy sample types and different modalities of analysis that can be performed to identify predictive biomarkers of RT treatment response

Fig. 7.6

Schematic of clinically informative elements obtained by liquid biopsy and various analysis methods. (Reproduced with permission from [40])

7.5.1.1 Tissue Biopsy

Tissue biopsies are the current gold standard for profiling tumors and can provide both key pathological and molecular information [41]. Numerous studies have investigated the potential of tumor tissue biopsies to predict the biological behavior of tumors, before and during RT, which could highlight the modes of biological action toward radioresistance. Examples of studies are provided in Table 7.4.

Table 7.4 Examples of studies that have used tumor biopsies to predict response to RT.

Immunohistochemistry (IHC) is a low-cost technique used by pathologists that involves staining fresh, frozen, or paraffin-embedded tissue and is widely applied in a clinical diagnostic setting. IHC has been used to identify predictive tissue biomarkers for RT response and include markers related to cell proliferation; ki67 and PKA (protein kinase), cell cycle checkpoint; p53 and p16, apoptosis; bcl2 and bax, growth factor receptors; EGFR (epidermal growth factor receptor) and finally, hypoxia; HIF1α (hypoxia-inducible factor 1-alpha (e.g., [42])). IHC facilitates the direct assessment of antigen expression in tissues through enzyme-conjugated antibodies. Initially, IHC was designed to classify the cellular origin of a tumor but with enzyme-conjugated antibodies and paraffin embedding, IHC is also capable of assessing treatment efficacy and is useful for tumor subtyping as well as in predicting patient response to RT. However, IHC is prone to pre-analytical subjectivity, operator subjectivity, and limited to known proteins [46]. Patient-derived 3D models can also be used to assess tumor response to RT and will be discussed in the following section.

7.5.1.2 Patient Tumor Tissue-Derived Organoids (PDOs)

In the last decade, patient-derived organoids (PDOs) have provided novel models for preclinical and translational research for assessing tumor response toward personalization of treatment. Organoids have the potential to be used as predictors for patient treatment response due to their ability to reflect the biological characteristics of primary tumors, i.e., intra-tumor heterogeneity, genotype, and phenotype [47] as well as the tumor microenvironment [48]. Compared to 2D models, PDOs possess improved cell morphology, differentiation, and viability, rendering them more relevant to the in vivo context [49]. Assays that can be performed on organoids include genomic profiling, survival assays, flow cytometric analysis, immunofluorescent, and histological staining. The main limitations associated with organoids include their high cost both in an economic and time-input sense [50].

7.5.1.3 Patient-Derived Xenografts (PDXs)

PDXs are mouse models that are widely used in modern cancer research and are proving to be another useful platform in the development of personalized medicine strategies due to an improved relationship with the context in vivo. PDXs also demonstrate similar susceptibility to anti-cancer therapies, they closely resemble patient tumor features, have similar histological and molecular characteristics, and can be cultured long-term in vitro [51, 52]. The tumor material to be used in PDXs is derived from fresh tumor tissue collected from a patient during surgery. Small tumor pieces are then implanted into severely immunodeficient or humanized mice. Although PDXs are a promising tool for translational research, they are difficult to apply as tumors may not grow or metastasize. Other disadvantages include the long process required to establish a model which requires significant involvement by pathologists, sampling and representational issues due to tumor heterogeneity, the overall economic cost of their development, their inability to evaluate the involvement of the immune system ex vivo, the potential for grafts to be rejected (“engraftment rate”), and the required use of regulated and approved animal facilities [53].

PDX models have been used to investigate biomarkers of RT response with the aim of stratifying patients based on risk and facilitating the individualization of treatment, as exemplified recently in PDX models of glioblastoma. The CHGA and MAPK8 gene signatures have been associated with increased survival in patients with glioblastoma who have received RT [54]. As the use of PDXs to reliably predict clinical activity of treatment options is still in its infancy, it is currently unknown whether these models can be used to guide individual treatment strategies in a time frame that is useful for a patient. Future technological advancements may accelerate their involvement clinically.

The invasive nature of tumor sample acquisition lends to many of the limitations associated with this sample type, including being painful and difficult to collect, time-taxing, having limited repeatability due to localized sampling of tissue. In addition serial assessments are often limited, and this diagnostic approach requires expert pathologists for evaluation, with the potential to introduce new risks to patients. Importantly a tumor may not be fully represented by a single tissue biopsy due to tumor heterogeneity which potentially adversely affects the accuracy of the test. Lastly, if the condition of a patient has worsened, the acquisition of tissue biopsy is not feasible [55, 56].

7.5.1.4 Liquid Biopsy

As the collection of tissue biopsies from patients often introduces unnecessary risks to the patient, there has been a recent increase in the focus on safer and less invasive sample collection methods, including via liquid biopsies. Liquid biopsies are generally a rich source of tumor-specific biomarkers, providing a temporal snapshot of the genomic character of a tumor, and can help overcome the complication of intra-tumor heterogeneity [56]. However, there are several limitations associated with liquid biopsies, including the lack of standardization of methodologies and inadequate technical/clinical validation for routine clinical utility [57].

It is known that intra-tumoral components are released into the bloodstream, urine, cerebrospinal fluids, pleural fluid, and so on, and that each contains information relating to tumor-specific material [58]. Blood is the most widely investigated liquid biopsy and where intra-tumoral components such as circulating tumor cells, circulating cell-free DNA, and extracellular vesicles (EVs) can be found [59]. These are the most investigated of the intra-tumoral components and will be discussed in this subsection (Fig. 7.6). Other intra-tumor components include circulating RNA, circulating proteins, and tumor-educated platelets; however, they will not be discussed in this subsection. Due to recent technological advances, these circulating biomarkers can be detected and researchers have identified them as a novel and promising avenue for stratifying patients based on risk and identifying patients who may be radiosensitive or possess radioresistant disease.

7.5.1.5 Circulating Tumor Cells (CTC)

CTCs have recently been discovered to be potential biomarkers for predicting tumor response to RT. A recent study by Qian et al. [60] demonstrated that nasopharyngeal patients with a complete response to concurrent chemoradiotherapy exhibited decreased CTC levels when compared to patients with a partial response. CTCs enter the bloodstream or lymphatic system and are disseminated throughout the body as they are released from primary, metastatic, or recurrent tumors. Metastatic tumors in distant locations can also form through CTCs that evade immune cell recognition [61]. These cells are rare, and the proportion present in peripheral blood is quite low when compared to white and red blood cells [62]. Molecular heterogeneity and the low concentration of CTCs in peripheral blood lead to multiple limitations in terms of their isolation, enumeration, and detection [63]. Current platforms to isolate and analyze CTCs are based on distinguishing features between CTCs and white and red blood cells such as morphology, biophysical and biomechanical properties along with modification, synthesis, regulation, and concentration of protein [64]. CTC isolation/enrichment platforms include microfiltration devices and dielectrophoretic field flow fractionation (DEP). CTC recognition platforms can be split into two groups: (1) label independent and (2) label dependent. The former includes PARSORTIX and CytoTrack. The latter includes iCHIP, CTC-Chip, and CELLSEARCH [64]. CELLSEARCH is an FDA-approved platform and is based on the expression of cell surface markers such as epithelial cell adhesion molecule (EpCAM), cluster of differentiation (CD)45 (CD45), cytokeratins 8, 18, and/or 19 [65]. This platform uses antibodies against these cell surface markers conjugated with magnetic nanoparticles or immobilized on microfluidic chips.

7.5.1.6 Extracellular Vesicles (EVs)

EVs can be isolated from a wide range of body fluids, such as bile, cerebrospinal fluid, saliva, breast milk, urine, blood, and amniotic fluid. Various cell-derived membrane structures are collectively termed EVs and include exosomes, microvesicles, and apoptotic bodies [66]. Exosomes are ideal candidates to study response to RT because radiation not only alters exosome manufacturing, but also affects their molecular cargo [67]. However, investigating the role of exosomes in radiosensitivity is a relatively novel approach, and studies currently are limited to in vitro studies that require translation in vivo to broaden our understanding of the mechanisms behind the development of radioresistance.

Current exosome isolation methods include immunoaffinity capture, ultracentrifugation, density gradient centrifugation, size exclusion chromatography, and exosome precipitation, while characterization methods include western blotting, ELISA, and transmission electron microscopy [68]. The disadvantages associated with these techniques currently make them unsuitable for clinical utility. These include (1) the large amount of starting sample and costly instrumentation that is required for analysis and (2) the labor- and time-intensive nature of the procedures required for sample isolation.

7.5.1.7 Cell-Free DNA (cfDNA)

cfDNA is reported to be found in elevated levels in cancer patients when compared to healthy individuals [69]. Usually, cfDNA is found in fragments ranging from 120 to 220 base pairs (or multiples thereof) [70]. The mechanisms responsible for the release of cfDNA into the bloodstream are not fully understood, but it is thought that it may be facilitated via apoptosis, necrosis, senescence, and actively through cell secretion [71]. In the blood, cfDNA is mostly nucleosome associated, and the tumor derived element in cancer patients is circulating tumor DNA (ctDNA) where concentrations of ctDNA have a linear relationship with tumor size and metastasis [72, 73]. Disease stage will also influence ctDNA concentration with late-stage disease associated with higher levels than early-stage disease [74].

ctDNA is more fragmented than cfDNA ranging from 100 to 200 base pairs and exists at much lower concentrations [75]. Detectable alterations that are tumor relevant include mutations, chromosomal rearrangements, copy number aberrations, methylation, DNA fragment lengths, tumor gene expression, and the presence of viral sequences (in tumors associated with oncogenic viruses) [73]. In patients with advanced stage nasopharyngeal carcinoma, plasma Epstein–Barr virus DNA load at the midpoint of RT is associated with a worse clinical outcome [76]. Detectable circulating HPV-DNA at the end of chemoradiation is associated with lower progression-free survival in HPV+ cervical cancer patients [77]. Somatic mutations in ATM, a DNA repair gene, can determine exceptional responses to RT in patients with head and neck squamous cell carcinoma, endometrial cancer, and lung cancer [78]. Several methods have been developed to extract and sequence ctDNA. These methods include, but are not limited to:

• 1. Polymerase chain reaction (PCR)-based techniques: BEAMing PCR (beads, emulsion, amplification, and magnetics), droplet digital PCR (ddPCR), and real-time quantitative PCR (qPCR).

• 2. Tumor-informed sequencing approaches: cancer personalized profiling by deep sequencing (CAPP-Seq), Signatera, and targeted digital sequencing (TARDIS).

Drawbacks associated with some of these techniques include that only one or a small number of mutations can be investigated at a time, a large amount of blood is needed to identify a small number of mutations due to low concentrations of ctDNA in the blood, and prior knowledge related to the tumor must be acquired before analysis, which usually requires invasive sample collection [79]. Next generation sequencing-based techniques include whole-exome sequencing (WES) and whole-genome sequencing (WGS) are also used for ctDNA analysis and limitations related to these techniques include that a large initial concentration of ctDNA is required and low sensitivity has previously been reported for WGS [80]. Examples of studies exploring these circulatory biomarkers can be found in Table 7.5.

Table 7.5 Examples of studies that have utilized biomarkers from liquid biopsies to predict tumor response to RT (see the referenced publications for the full study details, including RT regime)

Further studies are needed to elucidate the role of biomarkers in predicting tumor response to RT as currently there are limited studies that investigate their potential. Furthermore, biomarker identification is in its infancy, with liquid biopsies and 3D patient-derived models providing an enormous opportunity to further advance precision medicine. The limitations associated with these techniques may be mitigated in the coming years through technological advancements that allow the creation of more specific and sensitive assays. Along with harmonization and standardization of methodologies, the techniques mentioned in this section may move from a translational phase to routine clinical use. It can be expected that research on liquid biopsies and patient-derived 3D models will only grow in the coming years, and with this comes the potential to revolutionize patient care and treatment.

7.5.2 Predicting Normal Tissue Response

The radiosensitivity of normal cells, tissues, and tumors varies considerably between patients. There is variability in the patient’s response to RT, and most patients experience few or no side effects during or after treatment. However, due to this variability, many patients will receive suboptimal treatment dosing due to current dose thresholds being applied as a protective measure against toxicity events in radiosensitive patients. For patients who will develop side effects, a small number of these may develop more extreme side effects. Extreme side effects related to late radiation toxicity can be irreversible and life-threatening and greatly affect the quality of life of a patient. Identifying patients who possess intrinsic radiosensitivity prior to starting treatment would be clinically beneficial, as RT could do more harm than good in this small subset of patients. Identifying potential predictive biomarkers of normal tissue response to RT has been the focus of intense research within the clinical radiobiology arena over the years. Numerous attempts have been made by various research groups to develop an assay capable of predicting radiosensitivity, yet to date, no biomarkers to predict radiosensitivity are in clinical use. Assays have been developed with the aim of studying and predicting radiosensitivity in normal tissues and tumors. However, current developed methods have produced conflicting results and come with many limitations that make them impractical for clinical use (Table 7.6).

Table 7.6 Examples of studies using traditional and emerging techniques to assess radiosensitivity

7.5.2.1 Assessing Intrinsic Radiosensitivity

7.5.2.1.1 Cell-Based Assays

Cell viability assays are used predominantly to study cell response by measuring cell survival and proliferation after exposure to cytotoxic compounds. However, they are also extensively used in radiobiology studies. Clonogenic and MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assays are well-known assays for assessing in vitro radiosensitivity. The clonogenic assay is currently the gold standard for determining cellular radiosensitivity [96]. Further details on this assay can be found in Chaps. 3 and 7.

Early clonogenic studies provided evidence that in patients with cervical cancer and breast carcinoma, SF2 correlates with radiotherapeutic outcome [97, 98]. On the contrary, other early clonogenic studies have not found a correlation between SF2 and radiotherapeutic outcomes in head and neck cancer and multiforme glioblastoma multiforme [88, 99].

Numerous disadvantages are associated with the clonogenic assay, i.e., invasive sample acquisition, observer subjectivity through manual counting, merging of colonies that grow close together, long wait time for results as post-irradiation colonies can form 1–3 weeks later, labor intensive and technically difficult to perform [100,101,102]. As results from the clonogenic assay have a slow turn-around time, receiving results quite some time after sample collection/analysis would be of very little benefit to the patient, and it is clear that more efficient, rapid, and high-throughput methods need to be developed.

Plate-based cellular viability assays using tetrazolium salts such as MTT have also been used to assess radiosensitivity and can determine cell growth after irradiation [103].

Mitochondrial enzymes within metabolically active cells reduce the yellow MTT to water-insoluble purple formazan crystals. The amount of formazan crystals formed is directly proportional to the number of viable cells present in the sample, and this allows for the determination of viable cells through absorbance measurements obtained via a spectrometer at 492 nm ([103]; also discussed in Chap. 2). Buch et al. [103] also demonstrated that MTT performs similarly to the clonogenic assay when assessing the survival of irradiated human NSCLC and human glioblastoma cell lines. Compared to the clonogenic assay, the MTT assay is technically easier to perform and provides rapid results [104]. Rai et al. [105] also found that the MTT assay underestimated radiation induced cellular growth inhibition in numerous cell lines by comparing MTT values with cell numbers.

The MTT assay also has multiple disadvantages that make this assay unsuitable for routine clinical use, including:

• 1. Lack of specificity since tetrazolium reductions also reflect cell metabolism and not just cell proliferation.

• 2. Interference from reducing compounds.

• 3. Additionally, excessive direct light exposure of reagents and higher pH of culture medium can lead to sporadic reduction of tetrazolium salts resulting in raised background absorbance values.

• 4. MTT is cytotoxic and has been reported to inhibit cellular respiration leading to apoptosis [106,107,108].

7.5.2.1.2 Cytogenetic-Based Assays

These assays are used to identify chemicals and physical agents with genotoxic potential including irradiation.

During mitosis, MN are extranuclear bodies that are separated from the nucleus and contain defective chromosome fragments produced from DNA breakage and/or full chromosomes produced by interference of the mitotic machinery. MN and the micronucleus assay are discussed in detail in Chap. 3.

An early MN study carried out by Rached et al. [109] in the late 90s demonstrated that the MN assay had no predictive power for normal tissue reactions to irradiation. In this study, no variations in MN scores were observed between patients of various cancers who did or did not develop severe acute toxicities. Another study performed by Batar et al. [110] did not reveal any significant differences in MN scores between breast cancer patients who did or did not develop acute toxicities. However, a more recent study found a statistically significant difference in MN frequency per 1000 binucleated lymphocytes from patients who developed late cutaneous toxicity grade ≥3 when compared to grade ≤2 when irradiated with 10 Gy. Limitations of the MN assay include poor reproducibility due to high intra-individual variation and inter-laboratory variability, under certain conditions pseudo- MN can occur, and different types of chromosomal aberrations cannot be distinguished by micronuclei alone [111,112,113].

7.5.2.1.3 DNA Damage Assays

The γ-H2AX foci assay has also been explored as a prognostic technique for radiosensitivity. Please refer to Chap. 3, Sect. 3.6 (Cytogenetics and DNA Damage Measurements for Assessments of Radiation Effects) for more information. More recently, this assay was used to predict radiosensitivity in patients with oral squamous cell carcinoma and human colorectal cell lines [114, 115]. Other studies have also shown that γ-H2AX foci enumeration is not an ideal method for the prediction of acute and late toxicity development in prostate cancer, breast cancer, and rectal carcinoma [92, 116, 117]. On the contrary, other studies have used the γ-H2AX foci assay to identify patients at risk of developing radiation-induced toxicity in patients with lung cancer and breast cancer [93, 118]. These conflicting results on the fitness-of-use of γ-H2AX to predict intrinsic radiosensitivity further reinforce the idea that a novel method of analysis that can produce accurate and reproducible results needs to be developed. Disadvantages related to this assay include poor predictive performance and observer objectivity if γ-H2AX foci are enumerated by eye, and this is a fastidious and time-taxing process [119]. Inter-laboratory variations are also produced by this assay where significant variation in manually scored γ-H2AX foci yields obtained from irradiated lymphocytes has been observed [120].

The previously mentioned assays have limited clinical use due to their significant shortcomings, and a more practical approach needs to be developed to further investigate intrinsic radiosensitivity as a predictor of radiotherapeutic outcome.

7.5.2.1.4 Vibrational Spectroscopic Methods

Novel approaches to identify potential predictive biomarkers for radiosensitivity include Raman and Fourier transform infrared (FTIR) spectroscopic analysis of biofluids and cells. These techniques fall under the vibrational spectroscopy umbrella and are based on the transitions between quantized vibrational energy states of molecules due to the interaction between the sample and electromagnetic radiation [121]. Both techniques have numerous advantages over the previously mentioned predictive assays including minimal sample preparation and minimally invasive sample collection, speed, ease, and cost of analysis; they also allow for non-destructive and label-free analysis of a sample [121].

Each technique provides a biochemical fingerprint of a sample. Researchers in biomedical fields tend to focus on the range from 400 to 4000 cm⁻¹ and, in particular, the fingerprint region from 600 to 1800 cm⁻¹, as vibrations in these spectral regions produce refined bands and rich biochemical information related to disease prognostics and diagnostics. A major disadvantage of IR spectroscopy is the interference of water, which can overshadow crucial biochemical information [122]. However, Raman spectroscopy has a weak water signal and minimal water interference, making it ideal for the analysis of biological materials [123].

FTIR and Raman spectroscopy have recently been shown to be capable of discriminating patients on radiotherapeutic response [94, 95]. Both studies were successful in identifying variations in spectral intensities between patients with late toxicity grade 0–1 and grade 2+ with a high degree of sensitivity and specificity.

Current research from the same group involving Raman spectroscopy includes the analysis of biofluids and lymphocytes for the prediction of late normal tissue toxicity in high-risk localized prostate cancer patients and HPV+ head and neck cancer patients. Vibrational spectroscopy is far from being translated into the clinic, as currently there are no standardized protocols regarding the handling, storage, and preparation of samples to facilitate uniform spectroscopic analysis. However, researchers active in this area are currently investigating this to provide optimal protocols that will generate accurate and reproducible results [124, 125] (Fig. 7.7).

Fig. 7.7

Basic schematic of a Raman spectrometer

Raman spectroscopy is based on the inelastic scattering of light and scattering occurs when a sample is probed by a monochromatic light source. Most of the scattered light will be Rayleigh scattered where the laser photons will neither gain nor lose vibrational energy and will have the same energy as the incident light. Rayleigh scattering is also known as the elastic scattering of light and provides no information about molecular vibrational transitions [124]. A small fraction of light (approximately 1 in 10⁷ photons) is scattered at optical frequencies different from that of the incident light [125]. The Raman effect occurs when light probes a molecule and interacts with the electrons of the molecular bonds and the scattered light vibrational energy is not equal to that of the incident light. This process leaves the molecule in an altered vibrational state. Other light scattering processes take place with Rayleigh scattering, i.e., Raman scattering, and two forms exist: stokes (dashed arrow) and anti-stokes. Anti-Stokes scattering occurs when atoms or molecules lose energy during the transition from higher to lower vibrational energy states. Stokes scattering occurs when the atoms or molecules relax into a high vibrational excited state from the ground virtual state, resulting in a vibrational energy level higher than that of the incident light [124]. The Stokes scattered light will enter the Raman spectrometer and by a series of optics and mirrors will be directed to a monochromator. The collected light will be analyzed by the spectrometer and displayed as a Raman spectrum on a computer.

RT is a mainstay in cancer therapy, and due to recent technological advances, therapeutic efficacy has improved over the years. However identification of patients at risk of toxicity, as well as those who are radiosensitive or radioresistant remains a research challenge, which, if successful, could save patients from unnecessary treatment and avoid normal tissue toxicity and potentially result in improved tumor control. Numerous studies have attempted to identify a robust predictive biomarker of patient response to RT. Although the results of the studies have been promising, to date no biomarkers have been validated or translated successfully into the clinic. The assays mentioned here are not an exhaustive list of those currently being used for research purposes to study radiosensitivity (Box 7.4).

Box 7.4 Key Points in Relation to Use of Biomarkers in Clinical Settings

• There is patient variability in response to RT with most patients experiencing few or no side effects during or post-treatment.

• A small subset of patients may experience life changing and deliberating toxicities.

• Currently, no biomarker is in use in the clinic today to predict normal tissue toxicity.

• Disadvantages of conventional radiobiological assays deem them unsuitable for translation into the clinic.

• Novel methods such as vibrational spectroscopy demonstrate great potential in the hunt for a predictive biomarker.

7.5.2.2 G2 Chromosomal Radiosensitivity Assay

The G2 chromosomal radiosensitivity assay (also known as G2 assay) is a method that illustrates the existence of enhanced radiosensitivity and cancer predisposition based on the chromatid aberrations after G2-phase irradiation. For the evaluation of the individual radiosensitivity with this technique, peripheral blood lymphocytes are irradiated in vitro in their G2-phase of the cell cycle, incubated to allow repair of DNA damage, and blocked in mitosis by the use of colcemid, so that the chromatid aberrations can be observed and quantified. A high yield of chromatid breaks can indicate high radiosensitivity. This methodology has a major advantage as it enables a time-efficient individual radiosensitivity assessment.

The original G2 assay was developed by Sanford et al. [126]. However, a significant problem of this method was the high variability in radiation-induced damage observed in different samples even from the same donor. In addition, there was often an overlap between the G2 chromatid aberration yield in lymphocytes from healthy donors and cancer patients. Following further development, Terzoudi et al. [127, 128] proposed the use of caffeine in order to induce G2/M checkpoint abrogation, simulating this way the high radiosensitivity of AT patients. AT cells are known to have a defective G2/M checkpoint arrest and therefore AT patients are highly radiosensitive. With the use of caffeine, it was feasible to express the individual radiosensitivity in relation to the high radiosensitivity level observed in AT patients (Figs. 7.8 and 7.9). This protocol has a great advantage that it minimizes the effects of laboratory specific parameters and makes the inter-laboratory comparison feasible by enabling an ameliorated intra-experimental and inter-laboratory reproducibility. More recently, efforts have been realized for further optimization of the G2 assay by using other DNA Damage Response (DDR) and G2-checkpoint inhibitors—than caffeine—such as ATR- or ATM/ATR inhibitors (e.g., VE-821 and UCN-1). Of these inhibitors, VE-821 has been proven effective in a rapid radiosensitivity assessment of different cell lines as well as normal tissue and primary tumor cells [129].

Fig. 7.8

G2 chromosomal radiosensitivity assay. Chromatid breaks after 1 Gy of γ-irradiation as visualized at a metaphase peripheral blood lymphocyte from a healthy donor where four chromatid breaks are observed. (Reproduced with permission from [128])

Fig. 7.9

G2 chromosomal radiosensitivity assay. Chromatid breaks after 1 Gy of γ-irradiation as visualized at a metaphase peripheral blood lymphocyte after applying G2-checkpoint abrogation by means of caffeine where 13 chromatid breaks are visualized. (Reproduced with permission from [128])

7.6 Age-Related Radiation Sensitivity

Age at the time of radiation exposure is a key factor contributing for radiation-induced health effects, namely cancer. In a general way, it is accepted that individuals exposed at early ages are the most radiosensitive, whereas the latency period from primary damage to outbreak into cancer is longer. Then, radiation sensitivity decreases until maturity and increases again at older ages [130, 131]. In addition, considering that both cancer incidence and mortality rates increase with age, a model of radiation-induced cancer must also include the attained age. Attained age is defined as the sum of the age of the person at the time of radiation exposure and the period elapsed since the radiation exposure (“attained age” = “age-at-exposure” + “time since exposure”) [132].

Age-time patterns may also be represented as the “time since exposure,” corresponding to the “attained age” subtracting the “age-at-exposure.” For example, the cancer risk for someone with an age at exposure of 15 years and observed at an attained age of 40 years (time since exposure of 25 years) will be different from the risk for someone exposed at the same age but observed at an attained age of 79 years (time since exposure of 64 years) [132].

The understanding of the age-related alterations that may compromise individuals’ health after exposure to IR is increasingly relevant, along with the elucidation of the biological mechanisms underlying the aging-radiation exposure association. The fact that life expectancy of the worldwide population is steadily rising emphasizes the urgent need for a better understanding of the relationship between aging and sensitivity to radiation, which impacts radiation protection in clinical practice [130].

7.6.1 Epidemiological Evidence

Epidemiological studies developed in the Life Span Study (LSS) cohort of the Japanese atomic bomb survivors have provided valuable data on the relationship between the age at the time of exposure and oncogenic risks. The most standard models for radiation sensitivity, based on the measure of carcinogenic events, predict that the relative risks decrease monotonically with the increase of age at exposure, at all ages. However, new epidemiological data suggest that risks differ by age at the time of radiation exposure and by type of cancer (Fig. 7.10) [130].

Fig. 7.10

Schematic representation of the relationship between the relative susceptibility and the age at exposure. (Reproduced with permission from [133])

Data from the LSS cohort of the Japanese atomic bomb survivors showed that the excess relative risks (ERRs) of developing cancer following radiation exposure were higher during childhood and progressively decreased as a function of age until ages of 30–40 years old. However, for ages of exposure higher than 40, the ERR of developing solid cancer increased again. Thus, a bimodal distribution of radiation-induced cancer risks is associated with different biological processes. The greater susceptibility of children to radiation carcinogenesis is thought to be associated with three mechanisms: (1) long latency period between the primary injury and the cancer onset, that make children more likely to experience the long-term consequences, such as cancer; (2) faster radionuclides accumulation in growing bones compared to bones of an adult; (3) high frequency of cell division (as the one occurring in a growing organism) may allow an impairment of the radiation-induced DNA damage repair mechanisms. At a cellular level, this high radiosensitivity of children may be also related with the initiation of malignant processes due to the larger number of stem cells that can derive into cancer cells in younger people compared with aged ones. On the other hand, the radiation risks for individuals exposed at later ages are related to the age-related deterioration of cell functions, which can be responsible for an augmented susceptibility for oncogenic transformation [130, 131, 134].

The results of surveys targeting atomic bomb survivors also showed that the periods that relate to high radiation sensitivity vary according to the type of cancer. For individuals exposed while they were young, the risks of thyroid and stomach cancers and solid cancer as a whole are higher, while individuals exposed during puberty have an increased risk of breast cancer and people with 40 years old or older have increased risk of lung cancer [135, 136].

7.6.2 Mechanistic Interplay Between Age and Radiosensitivity

The higher radiosensitivity of individuals exposed at early ages is likely to have a long-term biological counterpart in their organisms. The mechanistic interplay between age and radiosensitivity is thought to be influenced by age-related cellular changes, such as impaired DNA damage repair, telomere erosion and accelerated cellular senescence, augmented susceptibility of cells to oxidative stress and inflammation, and radiation-induced epigenetic alterations (Fig. 7.11) [130, 131].

Fig. 7.11

Age-related cellular changes that may influence radiosensitivity and their mechanistic interplay. Compared to young cells, aged cells present increased impaired DNA damage repair, telomeres attrition, increased oxidative stress, and additional epigenetic alterations

Aged cells show a decline in the efficiency of the DNA damage response (DD) after radiation-induced DNA DSBs. The DDR should begin with the recruitment of proteins involved in both nonhomologous end joining (NHEJ) and homologous recombination (HR) repair pathways (see Chap. 3). Aged cells present several defects in these repair pathways such as delayed DDR kinetics, poor repair efficiency, and compromised repair due to chromatin reorganization as a result of aging. Thus, the aging process entails a disturbed nuclear organization that may compromise the recruitment of DDR proteins to the site, where they are needed, the nucleus. Irradiating aged cells will increase the damages in an already dysfunctional repair system, leading to irreversible damages. These damages are usually persistent and appear in a chronic mode increasing the damage burden in cells and tissues. Such accumulated damages can trigger enhanced inflammatory and immune system responses often leading to pathophysiological conditions like autoimmune disease and sensitivity to radiation and other types of environmental stresses [130, 137, 138].

Previous studies showed that telomere shortening relates to increased radiosensitivity. When aged cells escape from replicative senescence (a state of permanent growth arrest induced when shortened telomere length is attained), telomeres keep getting shorter, originating a greater number of uncapped chromosomes available to rearrangements. This loss of telomere integrity leads to an increase of genomic instability, which can initiate a carcinogenic process [130, 131].

Although there is some controversy about the causal relation between oxidative stress and an aged phenotype, it is known that aged cells have higher ROS levels production and a compromised antioxidant machinery compared with younger cells. This progressive loss of the pro-oxidant/antioxidant equilibrium compromises both cellular structures and homeostasis of aged cells. Exposing these aged cells to IR will unequivocally overload the antioxidant system, making them more susceptible to the IR-induced cell damages [139].

Radiation-induced oxidative stress can drive epigenetic alterations, such as (1) DNA hypomethylation through 8-OHdG methylation inhibition or DNA demethylation processes, (2) DNA hypermethylation through DNA methyltransferase up-regulation or DNA methylation catalysis, (3) histone modifications, and (4) miRNA expression. On the other hand, cumulative epigenetic alterations can also occur upon aging. Although this topic must be further explored, it seems to be a relation between epigenetic alterations and age-dependent radiosensitivity [131].

7.6.3 Clinical Perspective

In children/adolescents/young adults, RT remains essential for the curative treatment of brain tumors, Hodgkin lymphomas (HL), acute leukemias, Ewing and soft tissue sarcomas, neuroblastomas, nephroblastomas, or high-risk retinoblastomas. Life expectancy is long for the 80% of children/adolescents who are likely to be cured. The incidence of acute post-radiation complications, especially late, and proportional to the dose delivered may exceed that in adults. In fact, the cumulative incidence of overall iatrogenic sequelae 30 years after the end of treatment reaches more than 70% in this population compared to 15% maximum in the adult population after a median 3–5 years of follow-up. This late toxicity can lead to sometimes lethal sequelae with a major socio-economic impact (e.g., educational problems, parental mobilization, difficulties in entering the workforce, hospitalizations and costly symptomatic treatments and impoverishment, etc.). Along with high cure rates in this population, radiation-induced cancers appear with a probabilistic distribution (20-year cumulative incidence of secondary malignancies 3%)—but the incidence of which increases with the dose delivered and the duration of follow-up. Also note, however, that risk of secondary cancers increases in cancer patients who have not had prior radiation treatment as well as those who had chemotherapy [140, 141]. Differences are observed in the long-term, site-specific patterns of excess radiation inducing second malignancies between survivors of childhood cancer and adult-onset cancer, in terms of second malignancies histologic distribution, magnitude of risk, latency period, associated risk factors (genetic predisposition, environmental exposures, hormonal factors, and immune function) [142, 143].

On the other hand, RT is essential in the multidisciplinary management of the majority of cancer types in elderly patients, where it sometimes is considered as the first treatment option, often hypofractionated, and in others, as an alternative to surgery and/or chemotherapy.

In adults and the elderly, self-sustaining inflammation, fibrosis/atrophy, microcirculatory abnormalities are more readily associated with radiation-induced sequelae. In addition, in children, the manifestation of radiation-induced sequelae involves abnormalities in tissue maturation, delays (or even cessation) of growth of irradiated tissues—resulting in additional hypoplasia and/or hypofunction [144].

In adults, the distribution of individual radiosensitivity follows a Gaussian curve, and the toxicity observed in 5% of the most radiosensitive individuals is at the origin of dose recommendations to be applied to organs at risk in any patient in daily RT practice. There are many reasons to believe that this is not the case with children and seniors, with likely great variability with age. This is especially true if one consider all the changes in metabolic functions throughout growth and/or due to additional comorbidities, variations in tumor death, and tissue healing pathways with age and tumor predispositions associated with childhood cancer involving DNA repair pathways (assuming that this trait correlates with individual radiosensitivity).

Differences in organ development and tissue repair in children and adults have a significant effect on the expression of radiation injury. In many tissues, organ development is supported by cell proliferation from the prenatal period. During the development of each tissue, pluripotent embryonic stem cells differentiate into different unipotent lineages that will participate in mature tissue homeostasis. Some stem cells remain—ensuring self-renewal within the tissue. Thus, these two mechanisms are involved in growth as well as in tissue repair and regeneration. As the tissue matures, each organ thus contains a mosaic of dividing cells that are at rest either transiently or permanently. During childhood and adolescence, body tissues follow different growth patterns with their own kinetics. Not surprisingly, the rapid growth of normal tissues also seems to coincide with increased tissue radiosensitivity and, consequently, with a higher susceptibility to radiation-induced neoplasia. Overall, neurocognitive effects, development of muscles, and growth of bones are all sensitive to the age at treatment. For example, the intelligence quotient (IQ) deteriorates more in children irradiated on the brain before the age of 5 years compared to older children [145].

In contrast, in the elderly, the cells of the proliferative compartment move toward a permanent state of rest or senescence. This aging process becomes critical after injury because, although senescent cells are capable of metabolic functions, they lose their proliferative capacity after stress such as radiation. Therefore, late effects in the elderly population can be seen as the result of an interaction between their diminishing ability of cells to repair themselves after injury and their natural tendency to progress to a state of senescence, which itself can be accelerated by irradiation.

Only very rare cases of children have been studied for their individual radiosensitivity. Three genetic syndromes have formally been associated with RT-induced fatal non-cancer adverse tissue events: AT (homozygous ATM mutations), LIG4 syndrome (homozygous LIG4 mutations), Nijmegen’s syndrome (homozygous NBS1 mutations)—all characterized by the impairment in DSB repair and signaling pathways (see detailed information in Sect. 7.8). Noticeably, some syndromes also confer an increased individual cancer predisposition. Some cases of significant radiosensitivity with mutations of genes whose function was not expected in the radiation response have to be stressed—involving cytoplasmic functions or cell scaffold and membrane organization (for example, Huntingdon disease, Usher syndrome; [146]).

The individual radiosensitivity in children/adolescents and elderly is thus so far mainly unknown, and clinical trials are pending using individual radiosensitivity assays and very long-term observational studies (Box 7.5).

Box 7.5 The Influence and Impact of Radiosensitivity Related to Age and Aging

• Age at time of radiation exposure is a key factor contributing to radiation-induced health effects.

• New epidemiological data suggest that risk of developing cancer differs by age at the time of radiation exposure as well as by type of cancer.

• Cellular and molecular changes related to aging influence radiosensitivity.

• Differences in organ development and tissue repair in children and adults have a significant effect on the expression of radiation injury.

7.7 Biological Sex-Related Radiation Sensitivity

7.7.1 Introduction

The different individual radiosensitivity may be affected by genetic and individual factors as well as by lifestyle factors. One of the individual factors that have been associated with radiation sensitivity is biological sex.

7.7.2 Biological Sex Differences

The different radiosensitivity and radiosusceptibility for each sex could be due to biological factors. The human sex is determined by the sex chromosomes, namely by Y and X chromosomes. Whereas females have two X chromosomes, males have one Y and one X chromosome. These two sex chromosomes differ in terms of length, structures, and in the number and types of genes. Thus the X chromosome is approximately 160 Mb with more than 1000 genes while the Y chromosome is approximately 65 Mb with only 100 genes. The Y-specific genes are expressed mostly in testicular tissue with the SRY gene being the most important for determining male sex [147, 148].

Gene mutations or defects in gene expression of sex chromosomes dominantly affect the X chromosome and could be responsible for the death of the zygote and, consequently, decrease female birth. It also is hypothesized that differences in biological sex-related radiation responses could be due to immunological and hormonal differences apart from epigenetic and genetic factors. The mechanism underlying biological sex-related radiation sensitivity is still not clear and needs to be studied in more detail. Until now, it is known that the cellular response to radiation is highly complex and involves several processes like alterations in gene expression, signal transduction, repair process, and cell proliferation and death which could vary with sex.

7.7.3 Epidemiological Studies

Most of the data used for improving the knowledge on radiation-induced health effects and radioprotection was gained from the atomic bomb survivors in Hiroshima and Nagasaki (in Japan) and Chernobyl (in Ukraine). Also, the development of other epidemiological and cohort studies to evaluate the effects of radiation on the population exposed to high-dose levels due to medical reasons, radiation accidents, or from natural sources, contributed to the effects documented till now, as summarized in Figs. 7.12 and 7.13. The use of cancer risk models allows to predict the excess relative risk (ERR), i.e., the proportional increase in risk over the background absolute risk, and the excess absolute risk (EAR) of cancer, i.e., the additional risk above the background absolute risk, of cancer as a consequence of radiation exposure [134].

Fig. 7.12

Summary of biological sex-dependent health risks induced by radiation exposure. (Reproduced with permission from [149])

Fig. 7.13

Summary of most common morbidities induced by radiation exposure for each biological sex. (Reproduced with permission from [149])

The analyses of the several investigations done since 1945, based on data from Hiroshima and Nagasaki A-bomb survivors, revealed a higher incidence of cancer with elevated rated of leukemia, breast cancer, thyroid carcinoma, stomach and lung cancers, with risk for solid cancers varying with sex (Table 7.7). Related to radiation-induced lung cancer (LC), an increased radiation risk was evidenced and it is nearly four times greater for females than males [150]. Furthermore, a report on mortality from a follow-up from 1950 to 1997 showed that only ERR of cancer is far higher for females than males, without significance for EAR [134] (UNSCEAR 2000).

Table 7.7 Values for excessive relative risk per Sv (ERR/Sv) obtained from A-bomb survivors data for age at exposure of 30 years (UNSCEAR 2000)

Reports and studies based on the nuclear catastrophe in Chernobyl in 1986, where the population was exposed mostly to iodineIodine-131, showed an increased incidence of a range of cancers. Apart from the incidence of the same types of cancers report from A-bomb survivors (Table 7.7), there was also an increase in the incidence of bladder cancer and renal-cell carcinoma. There are still new cases of non-hematological cancers detected every year, so it is too early to present the final reports. The development of various radiation-related health problems in people living in the contaminated territories of Ukraine, Russia, and Belarus were more evident in women, inclusively by affecting their reproductive abilities and leading to an increase in the number of spontaneous miscarriages, mostly of the female fetus. Moreover, in Ukraine, the development of thyroid cancer was seen 2.5 times more often in females than in males who lived in contaminated territories [149, 151].

Data from the most recent nuclear accident in Fukushima is not yet enough to reach conclusions. Although the residing population was immediately evacuated from the most contaminated area, in the most affected areas that were not evacuated, average doses to adults in the first year were estimated to be <4 mSv, so discernable increases in related cancers are not expected (UNSCEAR 2021).

Studies from health care workers exposed to IR strongly indicate that occupational exposure leads to increased rates of IR-related cancers. Although these outcomes are associated with dose and also are age dependent, little attention has been given to biological sex [149].

A study carried out in Mayak workers about LC mortality revealed that ERR is four times higher for females than males, but the EAR is 0.43 less in the same comparison. Related to other cancers, the ERRs per Gy is also higher in females than males for lung, liver, and bone cancers. Moreover, a cohort study from Sweden, Denmark, and the USA about the carcinogenic effects of long-term internal exposure to alpha-particles radionuclides showed no significant differences between sex for solid cancer. The reduction of the female birth rate was also reported for the population living close to nuclear power plants or affected by nuclear testing [147, 149].

Thus, the epidemiological studies presenting separated risk coefficients for females and males do not present a consensus about sex-related radiation sensitivity. Although the studies from A-bomb survivors demonstrated higher ERR values for women than for men for all solid cancers, the corresponding EAR values are similar for males and females when sex-specific organs are not considered. The data collected from the radiation-induced occurrence of the same cancer type in different cohorts is also inconsistent. Although all these differences, the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) concludes that the EAR and ERR for total solid tumors are around two times higher in women than men, varying with site and organs, but for leukemia, no sex differences were observed [147, 149].

7.7.4 Animal Studies

Several experiments were carried out in animal models to provide clear information about the relationship between biological sex and radiation sensitivity as well as to confirm and expand the evidence obtained from epidemiological studies. Also, cell models were created from peripheral blood samples from healthy donors, cell lines, or primary cultures from organs of interest. These studies made it possible to study in depth the molecular mechanism inherent to sex differences in radiation sensitivity, namely to access cellular responses to IR, whole-genome screening for gene expression, and analyses of epigenetic regulatory mechanisms [147].

The data obtained through the analyses of gene expression as well as from epidemiological studies showed no correlations among the sex-specific expressed genes and corresponding cellular phenotypes. Nevertheless, a much higher incidence of thymic lymphoma and osteosarcoma have been found in female mice after treatment with 227Th than in male mice. Most recently, it was reported bystander effects in the non-radiated spleen of mice and rats varying according to the sex-specific differences. A sex-specific activation of distinct pathways was also suggested in mice, in response to whole-body irradiation as well as different tissues and organs irradiations with acute and chronic low doses [147, 149].

7.7.5 Differences in Radiation Therapy Outcomes According to Biological Sex

Box 7.6 The Influence and Impact of Radiosensitivity Related to Biological Sex

• The differences in radiosensitivity of males and females could be due to biological factors.

• UNSCEAR concludes that the EAR and ERR for solid tumors are around two times higher in females than in males.

• The identification of biological sex-related radiosensitivity will contribute to personalized dose and fractionation for RT as well as for radiation protection.

• Currently, the annual dose limits for occupational exposures recommended by ICRP do not recognize sex.

The severity of tissue reactions observed in cancer patients exposed to the same dose of IR during RT is assigned to differences in individual radiosensitivity [152]. It is generally believed that individual radiosensitivity is genetically determined based on the existence of certain hereditary diseases that we detailed more in Chap. 3. Moreover, the response to RT could also be influenced by other facts than the biological and physiological differences between males and females in organs and tissues, such as lifestyle, i.e., be sex related (Box 7.6).

There was evidence that females when exposed to the same whole-body exposure dose of IR have a greater risk of cancer than males [153]. Moreover, radiogenomic research has brought evidence that individual polymorphisms are correlated with treatment response with significant differences between females and males. However, the few available findings do not allow any conclusions to be drawn about sex-specific differences in radiation sensitivity, mostly because there are no studies that present enough data to support these hypotheses, without other confounding factors, or whose focus is specifically on sex [154]. So, the inclusion of biological sex or even gender as a variable in future randomized control trials or cohort studies will be crucial. The identification of the individual radiosensitivity of each patient will allow a personalized dose adjustment for RT as well as to use the values to improve the protection of occupationally exposed persons.

Regarding fractionation, high total doses delivered at a high-dose rate, in fractions, at appropriate intervals showed a lesser genetic effect in both males and females than the same dose delivered in a single fraction. It was also reported that the magnitude of its reduction is the same as the low dose rate effect.

7.7.6 International Commission on Radiological Protection (ICRP) Recommendations

Since its foundation, the International Commission on Radiological Protection (ICRP) has issued recommendations and guidelines related to the use of IR, based on the most recent scientific evidence and experience obtained through the years of implementation of the system of radiation protection. Until now, the annual dose limit for occupational exposure recommended by ICRP is not based on individual characteristics, such as biological sex or gender. Although the inclusion of these characteristics will increase the complexity of the model, it would be a critical step to overcome the actual generalized system used for determining the proper radiation protection for each specific case [134].

Two radiation research projects, Multidisciplinary European Low Dose Initiative (MELODI, https://melodi-online.eu/) and European Alliance Medical Radiation Protection Research (EURAMED, https://www.euramed.eu/), have been working on this topic defining the individual differences in radiation sensitivity as a key research priority.

7.8 Genetic Syndromes Associated with Radiation Sensitivity

7.8.1 Ataxia Telangiectasia (AT)

Ataxia Telangiectasia (AT) is an autosomal recessive neurodegenerative disease that is caused by mutations in AT mutated (ATM) gene. A-T was first described by Sillaba and Henner in 1926, however, its phenotypic spectrum was only expanded after the description of the ATM gene in 1995. A-T has a worldwide estimated prevalence of 1:40,000 to 1:100,000 and is related with a poor prognosis and a short life span, being chronic pulmonary diseases and malignancy the A-T-related most common causes of death [155, 156].

Fig. 7.14

Overview of DNA damage and repair pathways and most common genetic disorders

As the main known gene related to A-T clinical phenotype, ATM gene contains 66 exons and encodes the ATM protein, one of the three members of the PI3K-like family. ATM protein plays a pivotal role in the activation of cellular signaling pathways upon DSBs, apoptosis, and genotoxic stresses, such as IR. It functions essentially in the nuclear compartment; however, it is known that ATM is also present as a soluble protein in the cytoplasm [157]. In the nucleus, as part of DNA damage response upon DNA DSBs or oxidative stress, ATM is activated leading to a phosphorylation cascade of several target substrates involved in DNA repair, chromatin remodeling, cell cycle checkpoint, and transcription, namely P53 (S15), CHK2 (T68), and MDM2 (S395). In the cytoplasm, it is thought to be responsible for the functions of peroxisomes and mitochondria upon oxidative stress stimuli, as well as regulating angiogenesis, glucose metabolism, and telomere processing [155, 158].

A-T is a complex multisystem disorder characterized by a phenotypic heterogeneity, since patients show a broad range of clinical manifestations, including progressive cerebellar degeneration, immunodeficiency, oculocutaneous telangiectasia, increased metabolic diseases, radiosensitivity, and cancer predisposition. Other abnormalities can also be manifestations of A-T, such as dystonia, chorea, athetosis, tremor, and parkinsonism. Clinical heterogeneity of A-T can be assigned to different types of mutations that cause an impaired ATM protein expression or affect its function in different ways. Clinical and preclinical studies revealed that the presence of inactive ATM is more cancer prone and lethal than null ATM [156, 157]. A-T patients show potentially a 10–25% increased risk of developing cancer, due to their immunodeficiency. In childhood, the most common types of malignancy are leukemia and lymphoma, while adults may also develop different solid tumors, namely breast, gastric, liver, parotid gland, and esophageal carcinomas. It is described that heterozygous ATM mutations lead to an increased risk of 5.1 for the development of breast cancer, compared to the general population; while ATM monoallelic defects are associated with an estimated relative risk of ~3%. Although there are some controversy on the association between ATM mutations and breast cancer susceptibility, cancer screening guidelines are being developed for ATM-mutated carriers [155].

The first association between the A-T caused by ATM homozygous mutations and its higher human radiosensitivity was made in 1975. This hypothesis has been clearly strengthened over the years by the fact that several studies reported lower SF2 (survival fraction at 2 Gy) in ATM-mutated cells compared to other radiosensitive cases, reinforcing its higher radiosensitivity. These cells are inclusively characterized as hyper-radiosensitive (SF2 ranging from 1% to 10%). In a mechanistic point of view, it is proposed that ATM protein may act upstream of the molecular process of radiation response, namely upstream of the predominant DSB repair pathway, NHEJ. Although it is not yet fully understood which of the ATM functions has the biggest influence on radiosensitivity, the hyper-radiosensitivity of ATM-mutated cells is mainly explained by the deficient recognition of DSBs by NHEJ, as a consequence of the absence of an ATM kinase activity in the nucleus [158].

7.8.2 LIG4 Syndrome

DNA ligase IV deficiency or Ligase 4 (or LIG4) syndrome is an extremely rare autosomal recessive disease caused by mutations in DNA ligase IV. LIG4 was the first radiosensitive-severe combined immunodeficiency (RS-SCID) disorder to be described and belongs to the group of hereditary disorders associated with impaired DNA damage response mechanisms. Only few cases were recognized with LIG4 worldwide, the reason why its prevalence is difficult to estimate [159].

The LIG4 gene encodes a key component of the major DSB repair machinery, the NHEJ pathway. This pathway constitutes a multistep process that involves several proteins, such as Ku 70/80, DNA-PKcs, XRCC4, and DNA ligase IV, among others ([160]; see also Chap. 3). DNA ligase IV is an ATP-dependent ligase IV involved in the final step of NHEJ. It forms a complex with XRCC4 and then interacts with DNA-PKcs and XLF to rejoin a pair of DNA ends. DNA ligase IV also develops an important role in the production of T and B lymphocytes receptors, being recruited to repair programmed DNA DSB induced during lymphocyte receptor development [160].

All mutations of LIG4 gene identified in patients are located near its active site and are typically hypomorphic. This means they are not fully inactivating, since they do not affect ligase expression and maintain a residual but impaired activity of the enzyme (5–10% compared to the wild-type) [159, 160].

Clinically and morphologically, LIG4 syndrome is characterized by microcephaly, unusual facial features, growth retardation, and skin anomalies. Patients also manifest acute radiosensitivity, immunodeficiency, and bone marrow abnormalities. Some clinical phenotypes of LIG4 syndrome overlap with other genetic syndromes, like Seckel syndrome, NBS, and FA. Although the incidence of this disorder is very low, some patients were reported with malignancy, mainly lymphoma [7, 160].

The LIG4 syndrome is also considered a hyper-radiosensitive condition, which is caused by the loss of DNA ligase IV function and consequent impaired NHEJ activity, with a gross DSB repair defect. The first patient described with LIG4 syndrome developed acute lymphoblastic leukemia at age 14. The patient dramatically over-responded to cranial RT and died from radiation morbidity. Subsequent studies revealed a homozygous mutation in DNA ligase IV, which is located near the ATP binding site and is thought to hamper the formation of DNA ligase IV-adenylate complex, reducing its activity to ~10%. This post-RT fatal reaction made this genetic syndrome to be associated with hyper-radiosensitivity [7, 159].

7.8.3 Nijmegen Breakage Syndrome (NBS)

Nijmegen breakage syndrome (NBS) is a rare autosomal recessive disease mainly characterized by presenting microcephaly at birth, stunted growth, immunodeficiency, and high predisposition to cancer, without the manifestation of ataxia. NBS was first described in 1979 in a 10-year-old Dutch boy and, then, was formally reported in 1981 when a brother of the boy presented similar clinical features. NBS is estimated to have a prevalence at 1:100,000 live births worldwide, being most common in Eastern Europe [161]. NBS is a consequence of mutations in the NBS1 gene, also named NBN, on chromosome 8q21. It was determined a Slavic founder mutation, considering that most of the individuals with this syndrome are from Slavic regions and carry the same deleterious deletion, c.657del5. Eleven NBS-causing mutations have been identified, all of them in exons 6–10 of the NBS gene [162].

The NBN gene encodes a 754 amino acid protein named Nibrin (NBN), p95 or nbs1. Nibrin is part of the MRN (Mre11/Rad50/Nibrin) complex involved in the repair of DNA DSB, as well as in immune gene rearrangements, maintenance of telomeres, and meiotic recombination. When exposed to DNA damaging agents, the MRN complex is activated by ATM phosphorylation and localized to DNA damage sites forming protein foci at DNA breaks. Consequently, mutations in this NBN gene lead to impaired translocation of the Nibrin protein into the MRN complex impairing subsequent repair of the DNA DSB lesion [162].

The diagnosis of this syndrome is based on the identification of the main clinical manifestations and posterior confirmation by genetic analysis. The previous knowledge of disease-causing mutation in both alleles of the NBN gene allows the realization of prenatal molecular genetic diagnosis. NBS patients have a high predisposition to develop malignancies, being the syndrome with highest cancer incidence among all chromosomal instability syndromes. Till now, no specific therapies are defined, and the prognosis for NBN patients with malignancies is still poor [162].

The first documented case of radiation sensitivity observed in an NBS patient involved a 3-year-old microcephalic boy with medulloblastoma. Also, several in vitro studies have shown that NBS cells present high sensitivity to IR and radioresistant DNA synthesis. Thus, the NBS patients face several challenges in treating their presented malignancies, such as cancer, due to the limitation of using RT. In fact, considering the defective DNA repair system, the exposure of these individuals to radiation should be minimized and avoided when possible [161].

7.8.4 Xeroderma Pigmentosum (XP)

Xeroderma Pigmentosum (XP) is a rare hereditary autosomal recessive disorder with an incident rate of 1:250,000 in North America, and 1:1,000,000 in Europe, affecting both sexes equally [163]. XP is clinically characterized by the presence of pain induced by UV exposure, skin dryness, progressive pigmentary alterations, xerosis, several types of skin lesions and damage, and high incidence of malignant tumors affecting skin, head, and neck. In fact, acute severe sunburns are present in 50% of XP patients as a consequence of the hypersensitivity to sunlight. Some patients also showed neurological disorders and ophthalmologic degeneration [164, 165]. XP is caused by defects in seven complementation groups (XPA to XPG) which play a role in NER systems. XPC and XPA are the most prevalent in Southern Europe and North Africa. XPC is caused by mutation in the gene XPC, which contains 16 exons and is located in chromosome 3 (3p25), encoding for xeroderma pigmentosum group C (XPC) protein. The most frequent mutation in the XPC gene is a 2 bp deletion, c.1643_1644delTG, p.Val548AlafsX25 [164].

XPC is a protein with several functions in the NER system to repair DNA damage by recognizing the damaged bases and forming a stable complex with UV excision repair protein RAD23 (“HR23B”) protein needed for the recruitment of other actors involved in the removal of bulky DNA adducts. Mutation in these genes leads to an irreparable DNA damage that confers hypersensitivity to radiation to these patients, including UV exposure, and predisposition to develop malignancies [163, 164].

XP patients have 10,000-fold more probability in developing skin cancer than the general population. No cure is yet available for XP patients, so they need to be completely protected and isolated from any source of UV radiation. Although there is a correlation between XP syndrome and hypersensitivity to UV radiation, there are only few reports presenting the effects of using radiation therapy in XP patients with malignancies. Most of those reports did not show acute or chronic complication after treatment, probably due to the action of other repair pathways, such as NHEJ or HR instead of NER. However, there have been reported preclinical studies showing that some variants of XP could be more susceptible to IR and it is recommended that all XP patients should be classified before starting radiation therapy [166].

7.8.5 Fanconi Anemia (FA)

Fanconi anemia (FA) was firstly described by Guido Fanconi in 1927, a pediatrician who reported three children, brothers, with specific features: short stature, physical abnormalities, and anemia [167]. Defined as a rare genetic disease, FA is caused by pathogenic variants in at least 23 genes: FANCA, FANCB, FANCC, FANCD1/BRCA2, FANCD2, FANCE, FANCF, FANCG, FANCI, FANCJ/BRIP, FANC, FANCM, FANCN/PALB2, FANCO/RAD51C, FANCP/SLX4, FANCQ/ERCC4, FANCR/RAD51, FANCS/BRCA1, FANCT/UBE2T, FANCU/XRCC2, FANCV/REV7, FANCW/RFWD3, and FANCY/FAP100. All these genes play a critical role in DNA repair and genomic instability and can be organized in different complexes [168]. Classified as an inherited bone marrow failure syndrome, FA is the most common genetic cause of aplastic anemia and, besides that related to hematologic malignancies, is one of the most common genetic causes with a ratio of males to females 1.2:1 [169]. Concerning heritability, FA can be inherited in an autosomal recessive manner, an autosomal dominant manner (RAD51-related FA), or an X-linked manner (FANCB-related FA) [170].

This syndrome of impaired DNA repair and genomic instability, defined as complex and heterogeneous, is based on different mutations. FA patient cells are unable to perform different functions, namely repair DNA interstrand crosslinks, NER, translesion synthesis, and HR, inhibiting DNA replication and transcription, important cellular processes [171]. Related to IR, DNA damage and in particular DSB are the main alterations caused, with it reported that hypersensitivity to IR on FA mutation carriers translated not only into deterministic effects but also into stochastic effects [169].

FA is diagnosed at the median age of 7 years although symptomatic and asymptomatic family members have been described from birth to >50 years of age [172]. This syndrome is classified as multisystem disease, characterized by clinical features such as congenital malformations (short stature, skeletal malformations of the lower and/or upper limbs, abnormal skin pigmentation, microcephaly, and genitourinary tract and ophthalmic alterations), progressive bone marrow failure with pancytopenia presentation, typically presents in the first decade, often initiated with thrombocytopenia or leukopenia, and increased probability of hematologic (myelodysplastic syndrome or acute myeloid leukemia) and solid malignancies (head and neck, skin, and genitourinary tract; [170]).

7.8.6 Hereditary Breast and Cancer Syndrome

Hereditary Breast and Ovarian Cancer (HBOC) syndrome was first reported by the French physician Pierre Paul Broca, in 1866, when observed a greater predisposition to cancer in his wife’s family. This syndrome is an autosomal dominant disease, mostly caused by germline deleterious mutations in Breast Cancer gene 1 (BRCA1) and Breast Cancer gene 2 (BRCA2). The exact cancer risks depend on the type of pathogenic variant, being this syndrome mainly characterized by an increased predisposition to different types of cancer. These mutations affect all ethnic groups and races: in the general population, mutations in BRCA1 and BRCA2 genes are estimated to have a frequency between 1:400 and 1:500. However, in Ashkenazy Jewish people, the frequency of causal variants is higher: 1:40 [173, 174].

HBOC is mostly a consequence of mutations in BRCA1 gene, located in chromosome 17, and BRCA2 gene, located in chromosome 13. However, only 25% of cases are associated with these two genes. Therefore, other genes are associated with this syndrome and, currently, more than 25 genes have been associated, such as Checkpoint Kinase 2 (CHEK2) gene, AT Mutated (ATM) gene, and Partner And Localizer Of BRCA2 (PALB2) gene. Most of them encode proteins that, in conjunction with BRCA1 and BRCA2 genes, act on genome maintenance pathways. More than 1600 mutations in BRCA1 gene and more than 1800 in BRCA2 gene associated with tumor susceptibility have been described [173, 175].

BRCA1 and BRCA2 genes are tumor suppressor genes with a crucial role in the cell, since they encode proteins that repair DNA DSB through HR recombination, allowing the maintenance of genomic stability and tumor suppression. When exposed to IR, these proteins are activated, localize DNA damage, and repair it. In this way, mutations in these genes lead to an inefficient repair mechanism and to an increase in genomic instability, increasing the probability of cancer development [176].

The diagnosis of HBOC associated with mutations in BRCA genes is based on the identification of pathogenic variants in these genes through molecular genetic tests. HBOC patients have a high predisposition to develop different types of cancers, some of which at an earlier stage, such as breast cancer (in both sex) and ovarian cancer. Additionally, HBOC is also associated with an increased risk of developing prostate cancer, melanoma, and pancreatic cancer although to a lesser degree. Until the moment, there are no specific therapies defined, so early diagnosis in carriers of pathogenic variants in the BRCA1 and BRCA2 genes is crucial to apply effective surveillance and prophylaxis measures [173].

Since the twentieth century, several studies have been carried out to understand whether individuals with mutations in the BRCA genes are more sensitive to IR, trying to understand the role of exposure to IR in patients with HBOC and whether there are differences in the ability to repair of DNA damage between carriers and non-carriers of mutations in these genes. Although some studies show an association between the exposure of individuals with the syndrome to diagnostic doses and the development of cancer [177, 178], other studies fail to show any association [91, 179]. Thus, it is crucial to carry out more specific studies to obtain clear and objective conclusions in relation to this subject.

7.9 Toward Personalized Medicine: Future Perspective

Biological markers of changes in the body in response to radiation have long been used to assess radiation dose and exposure circumstances (see Chaps. 3 and 8). In recent years, with advances in technology and the sophistication of the markers, the potential to use biomarkers of the body’s response to radiation and other stressors to help predict treatment outcome and indeed to tailor treatments has begun to be explored [180].

Development, validation, and implementation of biomarkers are not a simple process (see Chap. 3). Firstly because our bodies are hugely complex systems relying on hundreds of thousands of changing and interacting processes at any one moment, many of which have associated measurable changes, there are a huge number of potential biomarkers based on the body’s complex response to IR confounded by a large number of other internal and external factors. Secondly, and just as importantly, there is huge variation in interindividual responses for most biomarkers. One recent study, for example, identified 40 blood-based biomarkers which could provide informative data on carbon metabolism, vitamin status, inflammation, and endothelial and renal function in cancer-free older adults alone [181]. Harlid et al. [182] also outlined the large number of potential biomarkers for risk predictive and diagnostic biomarkers for colorectal cancers, in a recent systematic review. In terms of molecular radiation epidemiology, a very large number of potential biomarkers have been identified, but despite a very large amount of work in this area, only one biomarker (based on transcriptional changes) has been identified as suitable to pursue now [14]. Furthermore, the practicalities of development of protocols and standard operating procedures for clinical use are also a barrier to implementation [183].

However, in wider clinical practice as well as for radiation medicine, biomarkers to support personalized intervention are in development and in some cases, already in use. For example, Karschnia et al. [184] reported improved survival in patients with advanced cancers of the central nervous system, following application of systemic targeted immunotherapeutic agents. Connor et al. [185] showed how implementation of novel image-based biomarkers to support RT has improved patient-specific therapy outcomes for glioma patients.

Going forward, despite the fact that the mechanisms of radiation resistance are still not well understood, use of miRNA in prostate cancer has shown promise. For example, Soares et al. [186] found 23 miRNAs which were involved in genetic regulation of prostate cancer cell response to RT. In the lung, Leiser et al. [187] recently demonstrated the potential utility of caveolin-1 (a membrane protein highly expressed in radiation resistant lung cancer cells) as a prognostic biomarker for response to treatment with radiation as well as for tumor progression, in support of precision medicine. And for cancers of the liver, De la Pinta [188] recently identified a number of candidate biomarkers of radiation response and toxicity and highlighted how close this field in particular is to use of such techniques to support personalized radiation medicine.

Indeed, use of large scale “omics” data together with machine learning or other artificial intelligence approaches has opened up a number of avenues of research. For example, Manem [189] compared five different machine learning based approaches in two existing radiogenomics datasets and found a large number of biomarkers associated with statistically significant pathways of response associated with surviving fractions of cells. New techniques in cellular barcoding are also proving incredibly interesting, with Wursthorn et al. [190], for example, recently demonstrating the use of this technique for assessing clonogenic survival in response to radiation and quantification of radiosensitivity as well as the contribution of stochastic and deterministic processes. Major bioinformatics studies can help in the identification of gene signatures as biomarkers for predicting normal tissue radiosensitivity. A key challenge is still the need for large scale, independent, validation of biomarkers in prediagnostic studies [182] as well as biomarker-driven randomized controlled trials [191]. Nevertheless, given the recent advances, use of radiation biomarkers to support precision radiation medicine is an exciting field in which large leaps forward are expected in a relatively short timescale.

7.10 Exercises and Self-Assessment

• Q1. Cite a genetic syndrome associated with both radiosensitivity and radiosusceptibility.

• Q2. Cite a genetic syndrome associated with radiosusceptibility, but not radiosensitivity.

• Q3. What is described as the attained age in relation to radiosensitivity?

  • (a) The age from birth to death

  • (b) The sum of the age at exposure and time since exposure

  • (c) The age between onset of cancer diagnosis and end of treatment

  • (d) The sum of the age at exposure and time of exposure

• Q4. What are the three mechanisms associated with the greater susceptibility of children to radiation carcinogenesis?

• Q5. Which of the following sentences are true or false?

  • (a) Aged cells may have a compromised repair due to chromatin reorganization.

  • (b) Senescence in older cells can be accelerated by irradiation.

  • (c) Abnormalities in tissue maturation is often seen as radiation-induced sequelae in adults.

• Q6. Is the radiation sensitivity higher for males or females?

• Q7. Why is sex important to consider in context of radiation therapy?

• Q8. Why is sex not yet included in the ICRP recommendations?

• Q9. Why is it necessary to identify biomarkers to predict the response of the tumor to radiation therapy?

• Q10. What circulatory biomarkers are of current interest in the field of radiation oncology?

• Q11. Why are liquid biopsies rapidly being adopted into translational research?

• Q12. What is the current gold standard for assessing radiosensitivity?

• Q13. What advantages do vibrational spectroscopic techniques have over conventional radiobiological assays?

• Q14. If normal tissue toxicity determines the total dose to be delivered to a patient, what outcome will this have on their treatment?

7.11 Exercise Solutions

• SQ1. Ataxia telangiectasia (AT) caused by homozygous ATM mutations is associated with fatal tissue reactions post-RT and high cancer proneness after exposure to radiation.

• SQ2. Li-Fraumeni’s syndrome (LFS) caused by heterozygous p53 mutations is associated with high cancer proneness after an exposure to radiation but LFS patients do not show adverse tissue reactions post-RT.

• SQ3. Alternative (b) is correct. (The sum of the age at exposure and time since exposure).

• SQ4. Long latency period between injury and cancer onset; faster radionuclides accumulation in growing bones; high frequency of cell division.

• SQ5. (a) true, (b) true, (c) false.

• SQ6. Females.

• SQ7. Consideration of the individual radiosensitivity of each patient will allow a personalized dose and fractionation adjustment for RT.

• SQ8. Due to the lack of scientific evidence to support the establishment of different annual dose limitations based on sex, as well as the complex social and societal issues associated with potential implementation of sex specific dose limits.

• SQ9. Identifying biomarkers of tumor response will allow stratification of patients based on risk and identifying patients who may not respond favorably to treatment. In turn, this will provide tailored and optimized treatment for patients.

• SQ10. Circulating tumor cells, circulating free DNA, and EVs.

• SQ11. Liquid biopsies overcome many limitations associated with tumor biopsies, such as minimally invasive sample acquisition, easy repeatability, lower cost, and a rich source of tumor-specific biomarkers.

• SQ12. The clonogenic assay is still the current gold standard for studying radiosensitivity.

• SQ13. Vibrational techniques involve minimally invasive sample collection, non-destructive, label free measurement of cells, and results can be produced in a short time frame.

• SQ14. Patients who are radiosensitive and undergo RT are at a higher risk of developing severe toxicity, and to circumvent this, the doses delivered to these patients will be at a lower dose than is necessary for adequate tumor control and a positive outcome of treatment.