Coherent Scatter

Low-energy photons can be briefly absorbed by the bound electrons of an atom. If the photon lacks the energy to remove the electron from the atom, the photon energy will be immediately reemitted as another photon. The reemitted photon has the same energy as the incident photon, and close to the same direction of travel. Because no energy is deposited, coherent scattering does not contribute to dose deposition, but the small deflections of the photons from coherent scatter can cause blurring in diagnostic images. Coherent scattering accounts for approximately 10% of interactions at 30 keV and is negligible for most therapeutic energy beams.

Photoelectric Effect

Photons having sufficient energy to ionize an atomic electron can undergo the photoelectric effect (Fig. 27-1). In this process, the photon energy is entirely absorbed. Some energy is lost to breaking the electron binding energy, and the rest is carried away as kinetic energy of the ejected electron. The probability of a photoelectric interaction scales with the cube of the atomic number (Z) and the inverse cube of the photon energy (E), making the photoelectric effect very sensitive to material type and much more prevalent for lower photon energies. The photoelectric effect is the dominant photon interaction in tissue below 30 keV.

Compton Scattering

When photon energy is significantly higher than the binding energy of an electron, the photon can scatter from the electron without being absorbed, as illustrated in Figure 27-1. The result of this interaction is a photon with reduced energy and new direction and a recoil electron with some fraction of the initial photon energy. The energy of the scattered electron varies with the scattering direction. An electron scattered in the direction of the incident photon claims most of the initial photon energy, whereas electrons scattered at greater angles have successively less energy. Compton scattering is only weakly dependent on Z and is the dominant photon interaction in tissue between 30 keV and 30 MeV.

Pair Production

Above 1.022 MeV, photons can interact in the presence of a strong nuclear field. The photon will disappear and spontaneously become an electron-positron pair (Fig. 27-1). The electron and positron will divide the initial photon energy between them to create their mass and kinetic energy. These particles will lose their energy as they interact with the surrounding materials. Upon losing all their energy, the electrons will be absorbed into an atom. The positron, on the other hand, will annihilate by interacting with a local electron, creating two 511 keV photons. (This annihilation reaction is what is detected during positron emission tomography scanning.) Pair production is the dominant atomic interaction in tissue for photons above 30 MeV and therefore has only a minor effect in radiation therapy, where energies are significantly lower.

Photodisintegration

Above a threshold energy, a photon can be absorbed into an atomic nucleus and cause one of the nucleons (a proton or a neutron) to be ejected. This process is called photodisintegration. Photodisintegration is more probable in high-Z materials (such as metals), and thus is more likely to happen during photon generation in a linear particle accelerator (linac) than in tissue. Neutrons produced in this manner can contribute a significant background radiation dose to patients receiving radiotherapy from very high energy machines. However, photodisintegration is negligible for accelerators operating below 10 MV.

Charged Particle Interactions

Charged particles will lose and transfer their energy to a medium through two mechanisms: collision and radiation. Collision energy loss by an energetic charged particle refers to the energy transfer resulting in ionization, excitation, and molecular damage. In a collision event, energy is absorbed in the medium at or very near the site of the interaction. Collision energy loss accounts for more than 95% of energy loss in tissue for therapeutic-energy electrons and is the major source of absorbed dose along the path of the electrons. Radiative energy loss occurs when particles are accelerated in the electric field of a nucleus and emit a fraction of their energy as a photon. This process, called bremsstrahlung, is relatively unimportant in tissue but is fundamental to the production of therapeutic photons in a linac.

Most electromagnetic interactions result from an interplay between photons and electrons, because many photon interactions result in atomic ionization and release of an energetic electron, with some of the electron energy converted back into photons through the bremsstrahlung process. Thus the effects of therapeutic beams passing through tissue can be described as a photon-electron shower, with the highly penetrating photons carrying energy deeper into tissue until a scattering event occurs, and the resulting scattered electrons depositing most of the resulting energy locally through collisional interactions.

Linear Accelerators

The most common modality used in radiation oncology is external beam radiation therapy. Although a small number of radiation therapy facilities generate external beams using radioactive sources such as cobalt-60, the vast majority of therapeutic electromagnetic radiation is generated in a linac. A linac is a device that accelerates charged particles (electrons) to velocities near the speed of light using oscillating electric fields to push the electrons through a series of accelerating cavities. A schematic of a linac is shown in Figure 27-2, A. Electrons are accelerated to energies typically between 4 and 18 MeV. Electric and magnetic fields focus and steer the high-energy electrons such that they strike a thin metal target that stops the electron beam, with some fraction of the electron energy converted to a spray of photons through the bremsstrahlung process. The bremsstrahlung photons, called x-rays, move approximately in the same direction as the electrons and have an energy spectrum, ranging from a few 10’s of keV up to the maximum energy of the initial electrons. The resulting photon beam then passes through a series of filters and beam-shaping elements that flatten and define the edges of the beam.

The dose from a photon beam is related to its intensity, defined as the number of photons per unit area. Two major effects serve to decrease the intensity of a photon beam as it passes through tissue. First, as with any photon source, the beam intensity decreases with increasing distance from the source, just as is the case for a light bulb. In addition, beam intensity decreases as photons are attenuated from the beam via various scattering and absorption effects. This process leads to a characteristic decrease in intensity versus depth that varies based on photon energy. Although the photon intensity begins decreasing immediately upon entering a material, the energy released through the photon interactions is spread over a few centimeters as the electrons scattered by the photons gradually lose their energy as they pass through the material. The resulting dose distribution is characterized by a region of rapid increase near the surface, a leveling off of dose at a depth of 1 to 3 cm, and a gradual dose falloff as depth increases. The plot of dose versus depth is called a percent depth dose curve, as shown in Figure 27-3. Because higher energy photons are more penetrating, higher energy beams will attenuate more slowly, leading to a more gradual decrease in dose with depth.

Linacs designed to produce photon beams can also be configured to produce therapeutic electron beams. Removing the photon-generating target and replacing it with a comparatively thinner electron scattering foil allows the transmission of the initial electron beam, but not without scattering the initially narrow beam into a broader distribution. Multiple filters and beam-shaping elements, as shown in Figure 27-2, B, produce an even distribution of customized shape at the surface of the patient. Electron beams lose their energy through different types of interactions than photons, leading to a different pattern of dose versus depth for electron beams. Rather than periodically removing photons from the beam through attenuation, electrons lose their energy gradually and at a relatively constant rate until the entire kinetic energy of the electron is expended and the particles simply stop. Ideally, this phenomenon would lead to a region of constant dose with increasing depth until the depth of full energy loss is reached, at which point the dose would drop abruptly to zero, with the depth of the dose drop-off being dependent on the initial electron energy. In practice, scattering and redirection of electrons within the beam lead to a mildly peaked dose plateau region and a somewhat more gradual dose falloff, as shown in Figure 27-3.

Radioactive Sources

Unstable isotopes can spontaneously decay to lower energy states, releasing energy in the process. This radioactive decay can result in the emission of therapeutically useful photons, electrons, or other decay products. The degree to which a sample is radioactive is called its activity and is defined as a number of decays per unit time. The activity of a sample depends both on how much of the isotope is present and how quickly the isotope decays. The historic unit of activity is the Curie (Ci), which is defined as 3.7 × 10¹⁰ atomic decays/second, corresponding to the decay rate of 1 g of radium-226. Activity is also specified in Becquerel (Bq), defined as 1 decay/second.

Because the rate of disintegration is proportional to the number of nuclei present, the absolute number of radioactive nuclei will decay exponentially. Activity is proportional to the number of nuclei, and thus a sample’s activity, and therefore its ability to deliver dose, will decay with the same exponential behavior. The decay is described by A(t) = A_oe^−λt, where A is the current activity, A_o is the activity at time zero, and λ is a decay constant for the isotope in question. The decay rates for various isotopes are more commonly given as the time required for half of the sample to decay away; this period is called the half-life of the isotope.

Therapeutically useful isotopes vary in half-life and in energy of emitted particles, as shown in Table 27-1. Isotopes emitting higher energy particles can deliver significant amounts of dose farther from the radioactive source than those emitting lower energy particles. One type of radioactive source, ⁶⁰Co, emits photons with an average energy of 1.25 MeV, which is sufficiently similar to photon energies found in linacs that ⁶⁰Co can be used as an external source to treat targets deep in a patient. Most other therapeutically useful radioactive sources emit lower energy radiation with less penetrating power and must be placed in close proximity to the area to be treated. Sources are formed into small sealed seeds, typically 1 to 5 mm in size, and can be inserted into the treatment area on a temporary or permanent basis. The dose rate falls off very quickly with distance from a seed, both because of rapid attenuation and the rapid spread of photons as they move away from the source.

Repair

The tenets of target theory suggested that the shoulder region of the radiation survival curve indicated that “hits” had to accumulate prior to cell killing. Elkind and Sutton⁵⁸ sought to better characterize how this damage accumulated and how the cell processed it. Even in the absence of detailed information about DNA damage and repair at that time—the structure of DNA had only been elucidated a few years earlier—a few facts could be gleaned. First, the hits that were part of the damage accumulation process, yet did not in and of themselves produce cell killing, were, by definition, sublethal. Further, this sublethal damage (SLD) only became lethal if and when it interacted with additional SLD. Elkind and Sutton conducted experiments that deliberately interfered with the damage accumulation process by delivering part of the intended radiation dose, inserting a radiation-free interval, and then delivering the remainder of the dose in what was called a “split-dose” experiment.

The overall surviving fraction of cells after a moderate to high radiation dose was higher if that dose was split into two fractions with a time interval in between than if it was delivered as a single dose. This phenomenon is illustrated graphically in the right panel of Figure 27-15. This finding suggested that the cells that survived the initial dose fraction had “repaired” some of the damage during the radiation-free interval. Therefore this damage was no longer available to interact with the damage inflicted by the second dose fraction, and thus a higher cell-surviving fraction resulted. Had the total dose been delivered as a single fraction, more damage in total would have accumulated, some SLD would have been converted to lethal damage, and the cell-surviving fraction would have been lower. The results of split-dose experiments turned out to be crucial to the understanding of why and how fractionated radiation therapy works the way it does, that is, that fractionation/protraction of a total radiation dose reduces its toxicity, in large part because of SLD repair. This phenomenon is termed the dose-fractionation or dose-rate effect, and it occurs after low LET radiation exposure in an undiminished manner regardless of the number of dose fractions.

When considering complete cell survival curves, SLD repair manifests itself as a return of the low-dose, shoulder region of the radiation survival curve. After an initial radiation dose and an adequate time interval for repair to occur, the response of surviving cells to graded additional doses is nearly identical to that obtained from cells without previous radiation exposure (illustrated in the left panel of Fig. 27-15).

Bedford and Hall59,60 generated in vitro survival curves for HeLa cells irradiated at various dose rates and found that the killing effectiveness per unit dose decreased as the dose rate decreased, to a point. A limit to this dose rate effect was reached; that is, further lowering of the dose rate did not produce a further decrease in toxicity. This finding is consistent with the idea that survival curves have negative initial slopes, as described by the “αD” component of the LQ survival curve equation. A further implication is that small differences in initial slopes of survival or dose-response curves for different tissues could be magnified into large differences when many small dose fractions or continuous low dose rates were used.

Repopulation

The sparing effects of fractionated external beam and brachytherapy are largely attributed to the repair of SLD; however, other factors can be involved as well. Repopulation is defined as a compensatory increase in cell proliferation in tissues in response to an injury that results in a large amount of cell killing. Stem cells (known or suspected) of normal tissues and tumors can begin to proliferate during and after a course of radiation therapy,⁶¹ although the mechanism(s) the tissue uses to affect this process and the time it takes to commence varies with the tissue. Some tissues, however, do not seem to be able to mount a repopulation response, or at least, not a prompt or robust one, presumably because they only possess very small numbers of stem cells, if any.

Repopulation is desirable in normal tissues because it facilitates the healing of common radiotherapy complications that can develop during or soon after treatment (e.g., oral mucositis in patients receiving radiotherapy for head and neck cancer). Repopulation of tumor cells, on the other hand, is quite undesirable because it has the net effect of counteracting the toxicity of ongoing radiation therapy. After radiation therapy is complete, repopulation also can lead to tumor recurrence.

In intact tissues, it can sometimes be difficult to tease apart the relative contributions of repair and repopulation to the overall dose rate effect, although generally speaking, repair-related effects occur over a time scale of hours to a day, whereas proliferative effects usually do not come into play for days or weeks (or more) after the start of radiotherapy. In normal tissues capable of mounting a proliferative response, molecular signals associated with the radiation injury seemingly need to reach a certain threshold before the repopulation begins in earnest, and for most tissues, this process takes a minimum of a week, and more commonly, several weeks. Tissues that exhibit a prompt and robust response to radiation injury (during or within a couple of months of treatment) and begin to repopulate are called “early responding,” and those that show a delayed proliferative response (more than 6 to 9 months after irradiation), if any, are referred to as “late responding.”

For tumors, especially carcinomas, the prevailing view (although it is not without its detractors) is that compensatory repopulation does not begin until approximately a month into the course of treatment.⁶² One clinical implication of this view is that overall treatment times should be kept as short as practically possible, because treatment time much longer than a month would allow for more tumor cell repopulation and therefore less treatment effectiveness. Some clinical data (mostly culled from fractionation studies involving treatment of head and neck cancers) suggest that up to a third of a typical 1.8 to 2.0 Gy daily dose fraction is wasted as a result of repopulation⁶² and that for each day of treatment beyond the time that repopulation begins, as much as 1% of local tumor control is lost.

Reoxygenation

Tumor blood vessels tend to be abnormal both structurally, functionally, and physiologically as a consequence of the dysregulated angiogenesis characteristic of tumors63,64 (see also Chapter 8). One consequence is that tumors can develop microregions relatively lacking in nutrients and oxygen. And although prolonged oxygen deprivation, that is, anoxia, will eventually kill even the hardiest of tumor cells, hypoxia—a state of very low tissue oxygenation on the order of 0.5% oxygen or less (corresponding to a partial pressure of oxygen of about 10 mm Hg)⁶⁵—is potentially survivable, particularly when it is intermittent or cyclic in nature. In contrast, the structure and function of the vasculature of normal tissues is such that it specifically avoids (in nearly all cases) the development of such gradients of oxygen and nutrients. Thus hypoxia has a built-in specificity for tumors, which, ironically, is an attractive feature clinically, where differences between normal tissues and tumors might be exploitable.

Unfortunately, as previously discussed, maintenance under hypoxic conditions renders cells more radiation resistant by upward of a factor of three (i.e., three times the total dose would be required to achieve the same biological end point). Thus a tumor containing even the smallest fraction of clonogenic, hypoxic cells could easily make or break the success of radiation therapy that is necessarily limited by the response of the aerobic, and therefore more radiosensitive, normal tissues.⁶⁶ Many types of experimental solid tumors in laboratory rodents and spontaneous tumors in veterinary patients and humans studied to date show an average pretreatment hypoxic fraction of approximately 15% to 20%, although the range of values is quite broad,^45,67 as might be expected given intra- and intertumor heterogeneity.

Of course, tumor hypoxia would not constitute a barrier to clinical success if, during the course of fractionated radiotherapy, the fraction of hypoxic cells decreased, ideally to zero. This process is called reoxygenation, and it has been studied extensively in experimental rodent tumors (e.g., see reference 68), and to a limited extent in human tumors as well (e.g., see reference 64).

The reoxygenation process can be either slow (days to weeks) or fast (hours), and it can be complete or incomplete, depending on the type and extent of hypoxia present in the tumor. Slower reoxygenation might be expected to occur in cases in which the type of hypoxia is chronic, such as might occur in cells that are multiple cell diameters away from vasculature and near or beyond the diffusion distance of oxygen (approximately 70 µm34,44). Physiologically speaking, the main ways that reoxygenation could occur under these circumstances is for aerobic cells closer to vasculature to decrease their oxygen consumption, which would occur if such cells are killed in large numbers by the radiation, or for the tumor mass as a whole to shrink in size, which would have the net effect of bringing the chronically hypoxic cells closer in to the existing vasculature. Some, although not all, human tumors shrink at least somewhat during the course of radiotherapy. Conversely, in cases in which hypoxia is intermittent or cyclic, reoxygenation could occur on the order of minutes to hours, as has been noted for many experimental rodent tumors.^69,70

Redistribution

Tumors that contain a sizeable fraction of actively proliferating cells will necessarily contain subsets of cells in different phases of their cell cycle, much more so than for some normal tissues in which cells are differentiated, no longer able to proliferate, and reside in G₀ phase. Furthermore, even for the normal tissues that do contain stem cell compartments consisting of undifferentiated, proliferating cells, these tend to be few in number compared with the tissue as a whole, nor do they necessarily proliferate rapidly. Thus tumors are much more heterogeneous than normal tissues with respect to cell cycle phase distribution.

It was the pioneering work of Terasima and Tolmach⁷¹ during the 1960s that led first to the development of a relatively simple method of obtaining populations of cells synchronized with respect to cell cycle phase, followed by the generation of radiation survival curves for each phase.

After a single dose of approximately 7 Gy of x-rays, Chinese hamster V79 cells in culture were noted to be most radioresistant in (late) S phase. Cells in G₁ phase were of intermediate radiosensitivity (i.e., more resistant at the beginning of the phase but somewhat more sensitive toward the end of the phase), and G₂ cells were increasingly sensitive as they moved toward the most radiosensitive M phase. This phenomenon was termed the cell cycle “age response” for ionizing radiation⁷¹ and is a general feature of mammalian cell radiation response.⁷² (The original work by Terasima and Tolmach^71,73 also showed this feature to be true for human HeLa cells.) Differences in radiosensitivity for cells in each phase were largely attributable to changes in the shoulder regions of their respective survival curves, with late S-phase cells characterized by broad survival curve shoulders, and mitotic cells showing strictly exponential cell killing with no survival curve shoulders (Fig. 27-16). It should also be noted that when high LET radiations such as neutrons are used in place of x-rays, the age response variation through the cell cycle is significantly reduced or eliminated.

Because of the age response through the cell cycle for low LET radiations, an initially asynchronous population of cells surviving a dose of radiation becomes enriched with more resistant cells (S-phase, and to a lesser extent, G₁-phase cells). This partial synchrony decays rapidly, however, in part because of radiation-induced perturbations in cell cycle progression and also because of natural variations in cell cycle phase durations. Such cells are said to have “redistributed,”⁷⁴ with the net effect of sensitizing the population as a whole to subsequent dose fractions compared with what would have been expected had the cells remained in their resistant phases. The redistribution process can be demonstrated in rapidly growing cells in vitro but tends to be dampened in vivo because of the longer (and dispersed) cell cycle times and lower fractions of actively proliferating cells in tumors. Attempts to take advantage of redistribution clinically, however, such as by carefully timing each dose fraction to correspond to times when tumor cells would be expected to be most radiosensitive, have been unsuccessful.⁷⁵ In fact, it has been difficult to demonstrate that redistribution has occurred in tumors (or normal tissues, for that matter) at all, other than by indirect inference, that is, that increasing resistance to each subsequent dose fraction during radiotherapy does not occur.

Sensitizers of Hypoxic Cells

The increased radiosensitivity of cells in the presence of oxygen is due to oxygen’s affinity for the electrons produced by the ionization of biomolecules. Molecules other than oxygen also have this chemical property, known as “electron affinity,”⁸¹ some of which are not consumed by the cell for other purposes, as is oxygen. Assuming that such an oxygen-mimetic compound is not otherwise used by the cell, it should diffuse further from capillaries and reach hypoxic regions of a tumor, sensitizing the hypoxic cells to radiation.

One class of compounds that represented a reasonable trade-off between efficiency as a radiosensitizer (i.e., compared with oxygen) and diffusion effectiveness was the nitroimidazoles, which include the drugs misonidazole, etanidazole, and nimorazole, all of which were the subjects of intensive preclinical and clinical studies over several decades. Unfortunately, the nitroimidazoles were found to cause adverse effects in normal tissues (in particular, peripheral neuropathy), which curtailed their clinical utility.82,83 In fact, the only one of this class of compounds currently approved for clinical use is nimorazole, which is used in Europe for subgroups of patients with advanced head and neck cancer.⁸⁴

Another class of hypoxic cell radiosensitizers does not sensitize cells per se but rather kills them outright. This action produces the net effect of (apparent) sensitization of the tumor because of the elimination of an otherwise radioresistant subpopulation of cells. Agents selectively toxic to hypoxic cells are termed “bioreductive drugs,” with the organic nitroxide tirapazamine (formerly SR 4233)34,85 being the lead compound.

In preclinical models, tirapazamine has been shown to outperform the nitroimidazole radiosensitizers using clinically relevant fractionated radiotherapy⁸⁶ and with some chemotherapy drugs either alone or in chemoradiotherapy regimens.⁸⁷ In humans, more than 20 clinical trials are ongoing or complete, including randomized phase 2 and 3 trials with tirapazamine combined with radiotherapy and/or chemotherapy (particularly, cisplatinum) for advanced head and neck, cervix, and lung cancers. Although tirapazamine has improved outcomes for some standard regimens, results overall have been mixed (e.g., see references 88 and 89). Thus although the drug has in no way become a staple of clinical practice, its activity against some tumors warrants further investigation, as does the search for more effective agents from the same or similar chemical classes.

Chemotherapy Drugs as Radiosensitizers

One of the major changes that has occurred in the clinical practice of radiation therapy during the past two decades is the large percentage of patients who are treated with concurrent radiation therapy and chemotherapy. This combination is used in two ways: chemotherapy used independently for its cytotoxic effect on the tumor, and chemotherapy used to enhance the effect of the radiation therapy through a drug sensitization effect. It is this second approach that will be discussed here. As mentioned earlier in this chapter, the value of radiation sensitization is dependent on producing a differential sensitization of the tumor compared with the adjacent normal tissue (therapeutic ratio). A drug has little value if its only effect is to decrease the amount of time the radiation therapy machine is turned on.

It has been known for decades that certain pharmaceutical agents can increase the tumor cell kill produced by radiation therapy.⁹⁴ One of the earlier approaches used in this manner, and one that is still very commonly used, is the combination of a fluoropyrimidine (typically 5-fluorouracil or, more recently, capecitabine) with radiation therapy for a variety of tumors, but especially those of the GI tract.⁹⁵ The underlying mechanism(s) of action of these fluorinated pyrimidines as radiation sensitizers is still not completely clear, but it most likely involves some combination of the drug’s selective toxicity to otherwise radioresistant S-phase cells, interference with DNA synthesis and repair, and ability to dysregulate the cell cycle.⁹⁴ At present a fluoropyrimidine is used routinely in the therapy of most tumors of the GI tract, including esophageal, gastric, pancreatic, rectal, and anal cancers. In the past, the combination had been used in head and neck and uterine cervix squamous cell carcinomas, although that practice is much less common at present. The details of these combinations are described more fully in the individual disease site chapters in this book. However, it should be noted that local control has been consistently improved by this approach and has had a beneficial effect on survival in some of these diseases.

Another drug commonly used with concurrent therapy is cisplatinum, an inorganic compound consisting of a coplanar complex capable of producing both inter- and intrastrand DNA and protein cross-links.⁹⁴ The cross-linking produced by cisplatinum most likely interferes with cellular repair of radiation damage, particularly DNA double-stranded breaks, resulting in radiation sensitization,⁹⁶ although as in the case of the fluoropyrimidines, additional mechanisms of action have also been proposed.⁹⁴ This combination is routinely used for head and neck and uterine cervix cancers (e.g., as described in references 97 and 98), with a major impact on long-term outcome. Mitomycin C is routinely used in the therapy of anal cancer,⁹⁹ temozolomide is in standard use in the therapy of high-grade gliomas,¹⁰⁰ and various other combination therapies have also been explored.

Molecularly Targeted Drugs and Biologics

Investigators have a great interest in the use of biological therapies in combination with radiation therapy. Although preclinical data have been generated for quite a few of these agents, with even newer compounds constantly being developed, the only two that have gone through substantial clinical testing and entered the treatment mainstream are anti-EGFR inhibitors used to treat squamous cell carcinomas of the head and neck¹⁷ and anti-VEGF therapies in the treatment of GI tract tumors.¹⁰¹

The use of cetuximab, a monoclonal antibody raised against the EGFR, has been shown to improve outcomes in head and neck cancers when combined with radiation therapy¹⁷ and is being explored in a variety of other clinical situations. Cetuximab competitively binds to the extracellular domain of EGFR, preventing activation of the receptor by endogenous ligands, which abrogates its near-constant signaling to cancer cells to proliferate. The antibody-receptor complex is then internalized and degraded, resulting in a downregulation of EGFR expression relative to the overexpression often displayed by diseases such as squamous cell cancers of the head and neck.

Bevacizumab exploits a different cancer hallmark, namely the ability of tumors to continuously recruit new blood vessels from surrounding normal tissues. This recruitment is accomplished through the production and release into their microenvironment by tumor cells of large quantities of (among other proangiogenic factors) VEGF, the principal stimulant of endothelial cells of the host vasculature to reproduce and migrate into the tumor mass, forming new—if immature and aberrant—blood vessels. The observation that tumors cannot grow beyond a size of approximately 2 mm³ without the support of neovascularization¹⁰² has led to accelerated development of angiogenesis-inhibiting agents, many of which target either VEGF or its endothelial cell surface receptor.^103,104

Bevacizumab is a humanized monoclonal antibody raised against VEGF.¹⁰⁵ The binding of the antibody to VEGF prevents it from binding to its receptor on the surface of vascular endothelial cells, which has the net effect of inhibiting the downstream signaling events that cause these cells to begin proliferating and forming vessel “sprouts” that migrate toward the tumor mass. At first blush, therapy that blocks new blood vessel formation might be expected to create more tumor hypoxia, and therefore more radiation resistance. However, findings of both preclinical and clinical studies with bevacizumab and related inhibitors of angiogenesis show that they most often promote radiation (and chemotherapy) sensitivity, not resistance. One hypothesis to explain this observation is that antiangiogenic therapy actually “normalizes” tumor vasculature¹⁰⁶ by selectively pruning away small, immature, abnormal tumor blood vessels, leaving the larger, most “normal-like” vessels behind. This process would lead to improved tumor oxygenation and better access of chemotherapy agents.

Bevacizumab had been shown to prolong both overall survival and progression-free survival in patients with advanced colorectal cancer when added to a standard chemotherapy regime consisting of irinotecan, 5-fluorouracil, and leucovorin.107,108 In persons with rectal cancer,¹⁰⁹ bevacizumab has been evaluated in combination with radiation therapy, and although initial reports were encouraging, with excellent early responses, it remains unclear whether the combination will improve clinical efficacy. Also, despite prolonged progression-free survival, no clear overall survival benefit of bevacizumab was seen in large clinical trials for NSCLC,¹¹⁰ renal cell carcinoma,¹¹¹ and metastatic breast cancer.¹¹² In response to the disappointing results in persons with metastatic breast cancer, the U.S. Food and Drug Administration has recently withdrawn its approval of bevacizumab as a first-line therapy, concluding that the drug’s benefit did not outweigh its adverse effects, which can include hypertension, proteinuria, bleeding episodes, and thrombotic events.¹¹³ Continued interest exists in trying to explore the vascular normalization hypothesis in a variety of tumor types.

Because many newer biological agents have been shown to have radiation-sensitizing properties in vitro and in animal systems, investigators have a great interest in translating this information to the clinic. For example, one class of drugs currently in clinical trials is the poly(adenosine diphosphate–ribose) polymerase or “PARP” inhibitors. The PARP family of proteins participate in a number of cellular functions, including detecting the presence of DNA strand breaks and activating signaling pathways that recruit and mobilize DNA repair proteins to the site of the damage.¹¹⁴ PARP1 is by far the most abundant member of this family and has been the most frequently targeted for inhibition. PARP inhibitors have been evaluated in preclinical models as potential radiation sensitizers, given that unrepaired or mis-rejoined DNA double-strand breaks are considered the lethal lesions with respect to the cytotoxic effects of radiation and that PARP inhibition might be expected to enhance this damage and increase toxicity.

DNA repair inhibition will only be clinically useful insofar as it is selective for tumors, because inhibiting repair in irradiated normal tissues could have adverse results. Fortuitously, PARP inhibition was found to be selective for tumors, in particular the tumors whose cells were already defective in at least one DNA repair pathway, as is the case for cells that carry mutations in the BRCA1 or BRCA2 genes, for example.115,116 Because cells from normal tissues have intact alternate or “salvage” pathways for DNA repair, the PARP inhibitors should be less toxic. Phases 1 and 2 clinical trials of PARP inhibitors are currently under way for brain tumors.

Tolerance Doses

Several important considerations in determining tolerance doses for different tissues include the radiosensitivity of the tissue’s component cells, the proliferative organization of the tissue (which determines the earliness or lateness of the tissue response to radiation), the volume of the tissue irradiated, and the tissue’s fractionation sensitivity. As such, it follows that tolerance doses for particular complications in particular tissues are necessarily average values that take all of these factors into consideration. That tolerance doses are average values is reinforced by the fact that these doses have been determined by pooling clinical outcomes data on thousands of patients over long periods (decades, generally). Bearing this in mind, several general findings about normal tissue tolerance are worth mentioning.

First, because tissues and organs contain, by definition, more than one cell type (each with its own inherent radiosensitivity), it follows that one tissue could manifest more than one complication after radiation therapy, with the severity of each determined by the radiosensitivity of the particular target cell whose death precipitates the complication and by the time-dose-fractionation schedule used. In addition, the severity of one complication does not necessarily predict for the severity of another complication, even within the same tissue. Using skin as an example, an early effect of treatment might consist of dry or moist desquamation (killing of basal cells of the epidermis), whereas late effects might include fibrosis (loss of tissue parenchymal cells and some dermal fibroblasts, resulting in overproliferation of survivors) or telangiectasia (damage to small blood vessels in the dermis, resulting in abnormal regrowth).

Second, although it is generally true that tissues known to be composed of radiosensitive cells have lower tolerance doses (e.g., the tolerance dose for a 5% risk of radiation-induced bone marrow aplasia within 5 years of treatment, the “TD_5/5,” is ≈2.5 Gy) compared with the tolerance doses for tissues containing more radioresistant cells (e.g., the TD_5/5 for rectal stenosis or fistula formation is ≈60 Gy),⁴⁵ this situation is not universally the case because radiosensitivity is not the sole determinant of radiation response. Early and late effects in irradiated normal tissues do result from the killing of critical target cells either directly or indirectly; however, the “process” that culminates in the complication is both complex and dynamic, involving radiation-inducible gene expression (including production of cytokines and growth factors), cellular signaling cascades, different modes of cell death, and compensatory proliferative responses.

Tolerance dose data for representative early- and late-responding normal tissues are shown in Table 27-2. Data are included for select normal tissues from three seminal publications (each about 20 years apart) that compiled such information based on exhaustive reviews of the literature and updated to reflect new knowledge about the etiology of normal tissue complications and changes in treatment technique over the years.

Radiation Carcinogenesis

With increasing numbers of long-term cancer survivors, one late normal tissue complication of radiotherapy that is of particular concern is the production of second malignancies. Leukemia accounts for about 20% of these second cancers, with the remainder presenting as solid tumors that develop in and around the previously irradiated site.117,118 Large epidemiological studies have assessed the breast and lung cancer risk in Hodgkin lymphoma survivors¹¹⁹ and leukemia and sarcomas in cervical cancer survivors.¹²⁰ Certain subpopulations of previously irradiated patients are at an even higher risk than the majority, including children and young adults (who would be expected to live long enough for such second cancers to develop, and, in the case of children, who are more radiosensitive than adults to begin with), immunocompromised individuals, and persons with a known genetic predisposition to cancer. Large studies of previously irradiated patients demonstrate that the risk of late cancer developing in an incidentally irradiated organ will very roughly double the baseline risk.¹²¹ Fortunately, the baseline risk in an individual organ is quite low, and thus the absolute impact of a radiation-induced cancer is also low, assuming there is clinical benefit in the original therapy. However, the impact of a second cancer occurring many years after a prior cancer developed can be severe.

Volume Effects

By convention, normal organs have been divided into what are called “serial” and “parallel” organs, by analogy with electrical circuits.¹²² A serial organ is one in which a severe injury at any anatomic point in the normal structure will produce a severe functional loss. The classic example of a serial organ is the spinal cord. If there is major damage to the cord at any level, from a small dosimetric hot spot, for example, all function can be lost below that level. Examples of parallel organs are the kidneys and the liver. Radiation injury to a portion of one of these organs will generally just decrease its function by an amount approximately equal to the percent of the organ that is destroyed but often will not produce a major functional deficit unless the irradiated volume is large.

Some organs behave as if they have both a serial and a parallel component. The liver can be viewed in this manner because the hepatic parenchyma functions as a parallel structure, but the biliary system near the hilum functions as a serial structure; obstruction of the common bile duct is a major complication affecting the entire organ. Thus the radiation oncologist can design radiation therapy fields that totally destroy portions of the parenchyma of the liver without a clinical problem to the patient, whereas damage to any tubular structure (the esophagus, for example) can produce severe consequences.

Fractionation Sensitivity

During the 1960s and 1970s, research interest among radiobiologists focused on the shape of the shoulder region of cell survival curves, the nature of dose-rate and dose-fractionation effects, and their relevance to the practice of radiation therapy. This research was timely given the growing appreciation clinically that tissue tolerances varied, sometimes widely, depending on the exact time, dose, and fractionation pattern used for treatment. This research ultimately led to the replacement of the target theory model of cell survival with the LQ model, as well as the use of the LQ model for a new, biologically based approach to clinical isoeffect modeling.

In multifraction experiments assessing skin reactions in mice, Douglas and Fowler¹²³ analyzed their data using what was then a novel method that allowed the derivation of values for the ratio α/β, parameters of the LQ formula. This new approach to isoeffect analysis focused on the shape of dose-response curves, particularly in the low-dose region, and by extension, that the critical parameter in radiotherapy planning was the size of the dose per fraction. More widespread use of this technique in subsequent years revealed that in most cases, a systematic difference was found between early- and late-responding normal tissues and tumors in their responses to different fractionation patterns. The α/β ratios tended to be low for late-responding normal tissue end points (on the order of 1 to 6 Gy, with an average of about 3 Gy), and high for early-responding normal tissue and tumor end points (usually 7 to 20 Gy, with an average of about 10 Gy). Table 27-3 shows α/β ratios determined for a variety of human normal tissues and tumors; worth noting are the few tumor types (melanoma and prostate cancer for example) that are exceptions to the general trend and are characterized by low α/β ratios.

Representative dose-response curve shapes for tissues characterized by low or high α/β ratios are shown in Figure 27-17. The curve with the low α/β ratio has a shallow initial slope at low doses but curves downward rather abruptly as the dose increases. Conversely, the curve with the high α/β ratio has a steeper initial slope and bends more gradually as the dose increases.

In parallel with the fractionation experiments used to determine tissue α/β ratios, other investigators¹²⁴ quantified the fractionation sensitivity of experimental tumors and normal tissues using a plot of the log of the total dose to achieve a particular tissue end point versus dose per fraction; this is called an isoeffect curve, and this method of plotting tolerance dose data had been in common practice in radiation oncology since the 1940s.¹²⁵ Plotted in this way, the curves for slowly proliferating or nonproliferating normal tissues such as the kidney and spinal cord, for example, were found to be steeper than those for more rapidly proliferating, early-responding tissues, such as skin and gut epithelium, and most tumors.^124,125 Taken together, the low α/β ratio and steep isoeffect curve suggest that late-responding tissues are more sensitive to changes in dose per fraction, thus experiencing greater sparing with decreasing fraction size than their early effects counterparts. This phenomenon is illustrated graphically in Figure 27-18. It was immediately evident that this phenomenon could be exploited for clinical benefit.

Simulation, Treatment Planning, and Delivery of External Beam Radiotherapy

After diagnosis, the first step in designing and delivering radiation therapy to a patient is called “simulation.” Simulation is a process for determining the proper selection and orientation of beams so that they properly overlap a target. Simulation requires the determination of patient dimensions for dose calculation and the determination and/or creation of identifiable reference points to ensure that beams are being aimed correctly. Historically, patient dimensions were determined by methods as simple as forming a lead wire or plaster cast around the patient. Target volumes and patient anatomy were identified and located by examining bony structures on a series of radiographs. These radiographs were generally acquired using a device called a simulator. A simulator is a diagnostic x-ray source configured to mimic the geometric beam delivery configuration of a radiotherapy linac. The x-ray source is mounted on a rotating gantry, similar to a linac, with a detector rotating in the opposite position on the far side of the patient. This approach allows fluoroscopic imaging from multiple angles, which facilitates positioning the patient so that the center of gantry rotation can be placed in the tumor area. Once this position, called the isocenter, is selected, radiographs are taken from all desired beam angles. These radiographs are taken as a record for comparison with future portal films (taken during treatment) and as an aid for the physician in refining each beam shape. Finally, room-mounted lasers indicate the location of isocenter on the patient’s skin, allowing these points to be marked by a small tattoo or similar marking. These markings can later be aligned with identically placed lasers in the linac treatment room to position the patient on each treatment day.

Poor soft tissue contrast from fluoroscopic imaging means that, on a conventional simulator, treatment parameters such as isocenter and beam boundaries can only be defined relative to visible landmarks, such as bony anatomy. More recently, techniques have been developed to use dedicated computed tomography (CT) scanners for “virtual” simulation, replacing conventional simulators in many radiation oncology departments. Virtual simulation is simulation that is based solely on the CT scan of a patient, and it is made possible by the ability to reproduce any arbitrary radiograph from a CT data set. In virtual simulation, the patient is placed in the orientation and position to be used for treatment delivery. A CT scan is acquired, and, with the patient still positioned on the scanning table, the physician uses the superior soft tissue contrast of the CT scan to select a location for isocenter within the target area (Fig. 27-19). An integrated laser system then moves to indicate the position of the physician-selected isocenter on the patient surface, allowing the external markers or tattoos to be placed for future alignment with the linear accelerator vault laser systems. The superior soft tissue contrast on CT images allows improved tumor localization and improved beam selection compared with fluoroscopy-based conventional simulators.

After simulation is completed, the CT scan and other images can be sent to a computer-based planning workstation, where the team of physicians, dosimetrists, and physicists can begin the treatment planning process. Treatment planning includes the identification of target volumes and normal structures requiring dose tracking and the selection and modification of beams to achieve specified dosimetric goals. The contouring process of identifying the boundaries of all treatment and avoidance structures takes place on the CT scan acquired at the time of simulation, but it may be supplemented by other imaging modalities, such as magnetic resonance imaging, positron emission tomography, or ultrasound. Normal structures are typically defined at their anatomic boundaries. The volume to be treated, defined as the target volume inInternational Commission on Radiation Units & Measurements Report 60 (ICRU 60),¹³⁴ is created by combining three structures: the gross tumor volume (GTV), the clinical tumor volume (CTV), and the planning target volume (PTV). The GTV includes all disease detectable on each imaging modality, including visible primary tumor and involved lymph nodes. The CTV is an expansion of the GTV that includes regions of possible or probable microscopic disease, including both adjacent tissue and draining lymph nodes. The PTV is a further expansion of the CTV to account for anatomic motion and expected variations of patient daily setup. Expansions from GTV to CTV to PTV vary with disease, technique, and patient history, but typical expansion numbers from CTV to PTV are usually on the order of 1 cm or less. Expansion from GTV to CTV is entirely dependent on the biology of the clinical situation. At times, no GTV is present (such as treatment in the postoperative setting), in which case the CTV is based entirely on the assumed high-risk areas for residual disease.

Once target and avoidance structures are identified and defined on the patient images, a specific strategy for dose delivery can be developed by the physicians, physicists, and dosimetrists involved. The essential task of treatment planning is to select, arrange, and characterize a group of radiation beams to deliver a high dose to the tumor while keeping the dose delivered to normal structures under acceptable limits. The treatment plan must be specified in terms of the total dose and the number of fractions in which the dose will be delivered. Individual beams can be customized by specifying the beam energy, orientation, and shape. Beams are typically shaped to match the cross section of the PTV plus some small margin to maximize overlap with the tumor while limiting the amount of normal tissue directly exposed to the radiation beam. This goal can be accomplished by designing a customized block made from an easily cast metal alloy such as Cerrobend, but on modern machines it is typically performed with use of automated beam-shaping devices such as a multileaf collimator (Fig. 27-20). Use of multiple beams overlapping on the tumor site can maximize the dose received by the tumor. The appropriate number, orientation, and relative weighting of beams in a treatment plan will depend on geometric factors, such as tumor location and the proximity of normal structures, as well as clinical factors, such as prior or planned surgery, overall patient health, history of prior irradiation, likelihood of future irradiation, and the effects of concurrent chemotherapy.

The dosimetry of a sample treatment plan must be evaluated for likely treatment efficacy of the tumor and the possibility of toxicity to normal structures. Dosimetry can be evaluated by examining isodose lines and dose-volume histograms. Isodose lines are a 3D representation of dose levels superimposed on the patient image and indicate regions of high and low dose on any user-selected plane. As shown in Figure 27-21, isodose lines correlate dose with anatomic location and can indicate regions of insufficient or excessive dose and suggest beam arrangements and weightings that can better meet the dosimetric goals for the plan. Aggregate dose to contoured structures are expressed as dose-volume histograms, which display the percentage or absolute volume of the structure receiving a specified dose or lower (Fig. 27-22). Most dose constraints applied to treatment planning are expressed in terms of dose-volume histograms. For example, a typical constraint imposed on treating lung lesions is to limit the total volume of lung receiving 20 Gy or more to 30%.¹³⁵ Determining appropriate dose constraints for normal structures from clinical data is critical; compilations of so-called “tolerance doses” for different normal tissue effects are generated periodically as clinical data accrues and have most recently been aggregated in the Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC) publications.¹³⁶

Intensity-Modulated Radiation Therapy

In conventional radiotherapy, treatment plans are designed to deliver a roughly homogeneous dose distribution to the tumor region. This goal is most easily accomplished by covering the target with overlapping homogeneous beams. For this reason, most modern linacs are designed to deliver just such flat dose distributions. Attaining a homogeneous dose distribution across a target can be complicated by a variety of factors. For example, geometric concerns such as beam obliquity or irregular patient surfaces commonly lead to uneven dose distributions across a given beam depth. In other cases, the presence of a sensitive normal structure in the vicinity of the target region may make it desirable that the photon intensity be lessened across all or part of a given beam. Varying the intensity of multiple radiation beams in a manner that improves the overall homogeneity of the dose to the target and helps to decrease the dose to normal tissue is referred to as intensity-modulated radiation therapy (IMRT).¹³⁷

The simplest form of intensity modulation, wedge filters, has been in use in radiotherapy since the 1960s. A wedge is an angled piece of high-density material that preferentially attenuates a photon beam on its thick end, resulting in a dose gradient across the beam. Wedges are often used to flatten the dose distribution for beams striking sloped surfaces, as illustrated in Figure 27-23. More recently, complex intensity distributions have been created by combining multiple beamlets of varying intensity into a single beam having an intensity profile customized to meet the needs of a specific patient (Fig. 27-24). The delivery of multiple beamlets for IMRT is made practical by the development of the multileaf collimator, which allows each beamlet to be formed and delivered in sequence without requiring the intervention of a radiation technologist. The widespread implementation of multileaf collimators has allowed IMRT to grow from an experimental procedure practiced at a small number of academic centers in the 1990s to a standard technique applied in clinics all around the world.

Calculating the proper intensity distributions for multiple intersecting beams is too complex a task for human hand calculations. Rather than having the physician and physicist create a treatment plan by manually selecting beam orientations and field shapes, IMRT fields are designed by computer algorithms called optimizers. The treatment planners specify the hoped-for dose distribution, expressed as dose goals and toxicity limits, and the optimizer uses mathematical techniques to alter the intensity profile of each beam until the dose goals are met to the greatest extent possible. The first optimizers only considered purely dosimetric goals, but recent efforts have been made to include biological end point models in optimization functions, such as effective uniform dose, tumor control probability, and normal tissue complication probability. IMRT optimization has been used in a variety of clinical circumstances and has been shown to have clinical benefits for treatments in many anatomic sites, such as treatments for prostate,¹³⁸ lung,¹³⁹ and head and neck cancers.¹⁴⁰

Once a treatment plan has been completed, approved, peer reviewed, and passed through a rigorous quality assurance program, the patient is ready to begin treatment. On each treatment day the patient must be positioned in an orientation identical to that used during simulation, and thus the position within the patient designated as the isocenter must be placed in the room position matching the center of rotation of the linear accelerator. Rough alignment can be accomplished by aligning the tattoos from the simulation step with in-room lasers in the accelerator vault. For more precise alignment of the patient, image-guided radiation therapy (IGRT) techniques can be used. Global patient positioning can be refined using video imaging of the patient surface¹⁴¹ or two-dimensional imaging of bony anatomy (MV portal films or kV radiographs). Techniques to image soft tissue structures to improve patient alignment include the use of in-room CT scanners, linac-mounted cone-beam CT scans,¹⁴² implantable fiducial markers detectable by radiograph, radiofrequency signal,¹⁴³ or ultrasound.¹⁴⁴ Adjustments to patient position based on these techniques can be made prior to radiation delivery. To ensure that the patient moves as little as possible between positioning and the delivery of radiation, several forms of patient immobilization may be used.

During the delivery of radiation, the patient lies as motionless as possible while the linac rotates around the patient from position to position, delivering the beams specified during the treatment planning. Treatments are delivered and supervised by radiation therapists, who operate the linac and monitor the patient and linac position to ensure that both are in the proper location. A typical treatment session lasts 10 to 15 minutes, although some cases involving extra imaging or delivery of high doses can last 1 hour or longer. The delivery of the radiation itself is undetectable by the patient, because the actual energy deposited in tissue for a typical 2-Gy fraction, although sufficient to cause irreparable cellular damage, will only raise the temperature of the tissue by approximately one one-thousandth of 1°F.

Simulation, Treatment Planning, and Delivery of Brachytherapy

The short range of radiation from most brachytherapy sources confines the majority of the dose to the immediate vicinity of the sources themselves, allowing highly conformal dose distributions through the careful placement of sources within or near the target volume. This arrangement allows a degree of sparing of nearby normal structures that may be superior to that achievable with use of external beam radiation therapy. Because of the high degree of conformity involved, brachytherapy is most effective for relatively small, well-localized tumors. Most radionuclides used for radiation therapy use γ-ray photon emission as the primary source of therapeutic energy. Early brachytherapy procedures used radium or radon sources, but these sources have been almost completely replaced by safer, artificially produced radionuclides such as cesium-137, iridium-192, iodine-125, and palladium-103 (Table 27-1). Brachytherapy sources are generally enclosed in multiple layers of inert material. The inert layers prevent leakage of the radioactive material and, for photon sources, absorb unwanted decay products such as α-particles and electrons. Brachytherapy doses can either be delivered over a short period (temporary implants) or gradually over the lifetime of the seeds (permanent implants).

Brachytherapy treatments are generally delivered using either interstitial or intracavitary techniques. In interstitial brachytherapy, radioactive seeds are surgically placed within the tumor volume. The seed placement can be permanent or temporary. Permanent seeds are deposited in the tumor volume using needles. Temporary seeds are placed and removed from the tumor volume through implanted catheters. A common permanent interstitial procedure is treatment for early-stage prostate cancer, in which approximately 100 ¹²⁵I seeds, each the size of a grain of rice, are placed in the prostate. Temporary interstitial procedures typically use fewer seeds, with MammoSite therapy for partial breast irradiation using a single ¹⁹²Ir seed applied through a catheter.¹⁴⁵

Intracavity treatments are the most common technique in current use for brachytherapy. In this form of brachytherapy, seeds are placed in existing body cavities in the vicinity of the tumor volume. Radioactive seeds are deployed in catheters arranged in specially designed applicators that are inserted into the appropriate cavity. Intracavitary treatment is most commonly used for treatment of gynecologic cancers but may also be used in other sites, such as the nasopharynx, the biliary tree, or the esophagus. Intracavitary brachytherapy treatments are always temporary, ranging in duration from a few minutes to several days.

Two additional brachytherapy techniques include surface application and intravascular brachytherapy. Surface application of brachytherapy sources is conceptually similar to intracavitary treatment, in that the radioactive sources are placed in an applicator and positioned adjacent to the treatment area, although in this case the sources are placed on an exterior surface. Surface application is common for treatments on the eye, including cancerous (choroidal melanoma) and noncancerous (pterygium) conditions. Intravascular brachytherapy deploys radioactive seed(s) through a catheter placed in a blood vessel. This technique has been used to treat tumors in the liver via the hepatic artery, where the radioactive particle is placed on special beads that are then embolized in the hepatic vasculature, effectively limiting the dose to the tumor area in the liver.

Stereotactic Radiosurgery and Stereotactic Body Radiotherapy Physics

As previously discussed, the long fractionation schemes used in external beam radiotherapy (e.g., 2-Gy doses delivered in 30 to 40 daily fractions) are designed to give normal tissue a chance to recover while still delivering a clinically useful dose to the malignant tissue. Although this approach helps minimize normal tissue toxicity, the goal of greater tumor control could be increased further by using large doses in fewer fractions (e.g., one to five fractions of 5 to 20 Gy each). The large doses used in hypofractionated treatments are highly toxic to normal tissues as well as tumors, and thus hypofractionated treatments can only be delivered safely and with acceptable normal tissue morbidity under the following conditions:

Stereotactic radiosurgery (SRS) refers to a single-fraction delivery of a high dose to an intracranial target. The term “radiosurgery” was coined in 1951 by a neurosurgeon named Lars Leskel,¹⁴⁶ who developed the original techniques for SRS in the 1940s using orthovoltage x-rays. Since that time, patients undergoing SRS have been treated with heavy charged particles, megavoltage x-rays, and γ-rays from radioactive sources. During the following decades, the use of SRS has been documented in more than 4000 publications,¹⁴⁷ demonstrating the effectiveness of the technique in treating primary and metastatic brain lesions, as well as nonmalignant functional disorders such as arteriovenous malformations and trigeminal neuralgias. The single-fraction doses in SRS are sufficient to kill any tissue, cancerous or otherwise, and thus great care must be taken to produce a high-dose region that is tightly conformal to the target area and to ensure that the dose is delivered to the correct location in the patient. Highly conformal dose distributions can be created using 100 to 200 overlapping beams (Fig. 27-25, A) or by delivering dose continuously as a linear accelerator moves around the patient in one or more arcs (Fig. 27-25, B). Positioning the patient correctly relative to the beams requires much stricter immobilization than is typically used for standard radiotherapy. Traditionally, patients would be placed in a rigid stereotactic head frame, which would then be positioned relative to the radiation beams by affixing the frame to either the treatment table or a room-mounted pedestal. The frame, which is affixed to the patient’s skull prior to patient imaging, provides a known offset between radiation isocenter and the tumor location, allowing dose delivery with an accuracy of approximately 1 mm. Newer SRS immobilization techniques have been able to approach the accuracy of stereotactic frames without requiring fixation to the patient’s skull, having the advantages of being less invasive and more reproducible if fractionation is warranted. Examples of these techniques include vacuum-based frame fixation to the hard palate and IGRT-intensive frameless radiosurgery. SRS can be delivered to multiple tumor sites in the brain during a single treatment and is generally limited to well-defined lesions 2 cm or smaller.

Stereotactic body radiotherapy (SBRT) is hypofractionated treatment applied to extracranial targets, such as in the abdominal, pelvic, and spinal regions. SBRT has a much shorter history than SRS, with the first clinical outcomes for SBRT reported in 1995.¹⁴⁸ Early SBRT studies focused on the treatment of lung and liver lesions, but more recent efforts have included studies of spine, prostate, kidney, pancreas, and gynecologic cancers. The delayed development of SBRT relative to SRS is largely due to increased difficulties in target localization. For example, whereas intracranial lesions can be expected to maintain a constant position relative to the patient’s skull, soft tissue lesions in the body can move relative to bony anatomy or any external markers used for patient setup. Minimizing this uncertainty requires the use of in-room imaging devices to ensure accurate beam delivery. Soft-tissue imaging may be facilitated by implanting radiographically visible fiducial markers in the region of the tumor. The extensive use of in-room imaging (IGRT) has largely replaced the use of stereotactic body frames in SBRT. Lesions in the thoracic or abdominal cavities may also move during beam delivery as a result of respiratory motion. Techniques for minimizing the effects of respiratory motion include:

All of these techniques increase the uncertainty of beam delivery, requiring larger margins around the GTV and consequently larger volumes of irradiated normal tissue. Partially because of this situation, SBRT treatments are typically delivered in three to five fractions.

Physics Considerations for Intraoperative Radiation Therapy

IORT treatment modalities require short-range radiation to minimize the dose to tissue behind the irradiated area, and include high dose rate brachytherapy sources (typically ¹⁹²Ir), low-energy photon beams, or electrons from linear accelerators. Early efforts in IORT required either a dedicated operating suite with permanent irradiation equipment or that the patient be transferred from the operating room to a shielded accelerator room. Most modern IORT in the United States is performed using mobile electron linacs. The advantages of electron linear accelerators include greater dose homogeneity over the targeted region, similar dose absorption in tissue and bone, and relatively modest shielding requirements. In a typical IORT treatment, the surgeon opens the patient and removes as much of the suspected cancerous tissue as possible. The electron linac is then maneuvered into position directly over the surgical site, and the tumor bed and other high-risk tissues are irradiated. The appropriate electron beam energy is selected based on the desired depth of treatment, and an applicator cone is attached to confine the beam to the intended area. The patient is then moved into position under the linac and the tumor bed and other high-risk tissues are aligned with the radiation beam. A dose of 10 to 20 Gy is then delivered to the treatment area, the patient is moved away from the accelerator, and the surgical team completes the surgery.

Protons

Protons are charged particles that can be accelerated and directed into tissue, where they deposit their dose. Protons deliver a lower dose to superficial tissues, then a higher radiation dose at depth where the tumor is, and then virtually no dose to normal tissues beyond the tumor, which provides an advantage compared with conventional x-rays, which deliver the highest doses (from a single beam) to more superficial tissues. Thus protons can allow for more precise delivery of the radiation dose to the tumor in comparison with normal tissues. The RBE for protons is approximately 1.1, and thus they are only slightly more biologically effective than x-rays or electrons. The principal advantage of proton therapy, therefore, is improved dose distributions.

Proton therapy is the most common type of heavy particle therapy. Proton beams are produced by accelerating ionized hydrogen in a cyclotron or synchrotron to energies in excess of 100 MeV. These devices are considerably larger and more expensive than conventional clinical accelerators, and thus they tend to be built as stand-alone facilities, although research into smaller scale proton devices is ongoing. Unlike electrons that lose energy roughly evenly across the range of therapeutic energies, protons lose their energy at an increasing rate as the beams loses energy with depth. This effect culminates in a region of rapid dose deposition near the depth of maximum penetration, called the Bragg peak (Fig. 27-26). The depth of the Bragg peak increases with beam energy, allowing careful energy selection to ensure that the highest dose is delivered to the tumor volume, with lower doses upstream of the target and negligible dose downstream of the target. The width of the Bragg peak is virtually always narrower than the region to be targeted, requiring a range of proton energies to “paint” high, uniform doses across the target. This widened dose distribution, known as a spread out Bragg peak, can be created either by varying the proton beam energy in a synchrotron or by using a spinning modulator of varying thickness to selectively vary the beam energy, shown in Figure 27-26.

The largest clinical experience with the use of proton radiotherapy has been for the treatment of prostate cancer. However, although their use in this site has a theoretical advantage, that advantage has never been demonstrated in clinical trials.¹⁵⁰ Protons are very appealing for the treatment of certain pediatric tumors, because the effects of unwanted irradiation of normal tissues can be severe. In this context, protons have been used extensively in the treatment of pediatric central nervous system tumors.¹⁵¹ Studies exploring the use of protons for therapy of lung cancer and tumors at other anatomic sites are ongoing. Because the cost of a proton irradiator (and its specialized facility) is far greater than for traditional x-ray and electron linacs, it is incumbent upon persons using protons for radiotherapy to demonstrate that the extra expense translates into better tumor control rates and/or fewer normal tissue complications. The lack of such level I evidence to date has made many persons concerned about the rapid proliferation of proton facilities.

Neutrons

Neutrons are heavy particles that are not charged, and they too are produced by specialized machines. Neutrons interact with matter through different mechanisms than the photons and particles previously discussed; these interactions can cause low-energy protons and heavier ions to be ejected during collisions between neutrons and target nuclei. These ejected particles cause biological damage in accordance with their LET. Neutron beams therefore have an energy-dependent, average RBE that can vary between 5 and 20.

Neutrons have no dose distribution advantage over x-rays and in fact have a disadvantage in dose delivery. However, they are more potent biologically because of their higher RBE and therefore have the potential to destroy more effectively certain tumors that are relatively radioresistant, including melanomas, sarcomas, and salivary gland tumors. Many clinical studies involving neutron therapy have been conducted, but it has been difficult to find a clear advantage to this approach over conventional treatment methods.

Heavy Ions

Investigators have had a long-standing interest in using other heavy charged particles, such as accelerated carbon or neon ions, for therapeutic purposes, and that interest continues to the present day. These particles have the potential to combine both the biological advantages of neutrons and the dose distribution advantage of protons. Machines to produce these particles and the associated operating costs are very expensive, however, which has substantially limited their development and use to only a handful of facilities worldwide.152,153