Interaction of Chemotherapy and Radiation

Published on 03/04/2015 by admin

Filed under Hematology, Oncology and Palliative Medicine

Last modified 22/04/2025

Print this page

This article have been viewed 4467 times

Chapter 4 Interaction of Chemotherapy and Radiation

Oncology has increasingly become a multidisciplinary field of medicine: (1) surgery remains the definitive local treatment modality; (2) chemotherapy remains the definitive systemic treatment modality; and (3) radiation therapy is the definitive locoregional treatment modality. Although in the past these approaches were predominantly used exclusive of one another, in the past 20 years there has been an explosion of preclinical and clinical efforts to combine these therapies for improved outcomes, including improved locoregional control, overall survival rates, cosmesis, and organ preservation. Researchers have learned a great deal about the interactions between chemotherapy and irradiation from clinical trials that combined these treatment modalities in sequential and concomitant regimens. In addition, laboratory investigations have demonstrated key molecular targets and pathways that can potentially be exploited for improved outcomes. The combination of chemotherapy and irradiation has changed the management approach in several disease sites that are broadly reviewed here.

Historical Perspective

Radiosensitization and chemosensitization are each complex concepts that have many different interpretations and have been used to describe a variety of interactions.1,2 The combination of irradiation and chemotherapy for mutual or even simultaneous sensitization increases the intricacies of these interactions. For example, over 100 years ago radiation treatment and benzene systemic therapy were combined for leukemia treatment.³ However, probably the best historical model of chemotherapy and radiation therapy interaction is the use of 5-fluorouracil.

5-Fluorouracil

In the 1950s, the halogenated pyrimidine 5-fluorouracil (5-FU) was combined with external beam irradiation therapy (EBRT) after this class of drug was determined to have anticancer properties.⁴ In the past 50 years, 5-FU has been successfully combined with radiation to treat a variety of gastrointestinal cancers, as well as cervical cancer and head and neck cancer.⁵ The route of administration and scheduling of 5-FU doses have been reconfigured many times to reduce toxicity and maximize tumor control. What began as bolus delivery at the beginning and end of a fractionated radiation treatment course (Moertel regimen) has progressed to protracted venous infusion (PVI), and now to twice-daily doses of oral 5-FU analog formulations. These approaches have allowed for an increase in cumulative dose of the drug, decreased chemotherapy toxicity, and improved radiosensitization. 5-FU has proven to be a staple drug in the armamentarium of medical oncologists as well as a key radiosensitizer for the radiation oncologist.

Rationale

Limitations in Current Therapeutic Approach

Over the past several decades, there have been major technologic advances in surgery and radiation as well as rapid development of novel systemic agents. Nevertheless, cancer morbidity and mortality remain major problems. Combined-modality therapy seeks to improve on the limitations that surgery, chemotherapy, and radiation therapy carry independently. Radiation complements surgery by improving locoregional control, but tumor-specific and patient-specific factors limit the success of both surgery and irradiation. This chapter focuses on how systemic therapies are used to overcome the shortcomings of radiation treatment, such as the presence of micrometastatic disease, disease outside of the treatment field, and the inability to deliver an adequate dose to the target region because of the risk of toxicity to normal tissue. In addition, tumors may contain regions of relative hypoxia or subpopulations of cells with intrinsic or acquired resistance to radiation damage. A brief review of the current understanding of these topics follows.

Tumor Detection

If ionizing radiation were not toxic to normal tissues, tumor detection would be immaterial and radiotherapy could be delivered to the entire body much as chemotherapy is given. Obviously, this is not the case. Just as surgeons must identify tumor tissue and remove it with a scalpel, radiation oncologists must be able to identify tumor tissue so that they can precisely and accurately target it with radiation. Fortunately, advances in radiology have dramatically improved the ability to detect tumor location and extent. CT and MRI scanning provide excellent anatomic information, but when they are combined with biologic or functional imaging techniques such as positron emission tomography (PET) or single-photon emission computed tomography (SPECT), they are even more precise at differentiating tumor tissue from nontumor tissue. Emerging MRI sequences, including dynamic contrast-enhanced MRI or CT scanning, fast imaging employing steady-state acquisition (FIESTA), and ultrafast pulse sequence approaches, provide improved anatomic imaging for surgeons and radiation oncologists. Nevertheless, the resolution of current techniques (on the order of 5 mm for PET resolution ⁶) and the high rate of false-negatives with small tumors still limit the ability to identify microscopic tumor extent and micrometastatic disease. Technologies currently being explored, such as PET/MRI, as well as novel radiopharmaceuticals may provide improvements in the near future.

Inherent and Acquired Resistance

Certain tumors have inherent radiation resistance pathways that manifest with high rates of local failure following irradiation. In some cases, such as pancreatic cancer, it is difficult to deliver adequate doses of radiation to the target because of the limitations in dose tolerance of surrounding bowel, kidney, and liver tissues. Other tumors have extremely high local failure rates despite dose escalation; for example, glioblastoma multiforme has local recurrence rates approaching 100%. Biologic factors within the tumor or tumor microenvironment also generate resistance. Figure 4-1 summarizes some of the major resistance pathways within tumors. Later in this chapter we will describe how chemotherapeutic agents can potentially mitigate these resistance pathways.

Figure 4-1 Schematic of radiation-induced signal transduction cascades indicating pathways leading to anti-apoptosis, proliferation, DNA repair, and angiogenesis.

Increased Toxicity

In the original treatise by Steel and Peckham ⁷ on combining chemotherapy and radiation therapy, it was assumed that each modality functioned independently in both beneficence and toxicity. However, concurrent chemoradiation therapy has shown increased toxicity, indicating some level of overlapping toxicity, chemosensitization by radiation, or radiosensitization by chemotherapeutic agents. Because chemoradiation is often used when tumors have wide anatomic extension (and thus surgery is precluded), the volume of normal tissue irradiated—and at risk for toxicity—is larger. In some cases, a patient has co-morbid conditions that prevent aggressive therapy.

Therapeutic Index

A metric is necessary to determine the efficacy of a therapy relative to its toxicity, so that newer approaches can be compared. This metric, known as the therapeutic index (or therapeutic ratio), is the ratio of the probability of tumor control to the probability of normal tissue toxicity. Typically, the ratio is calculated based on the 50% control rate of tumor tissue versus the 50% rate of normal tissue toxicity. These sigmoid-shaped curves determine estimated efficacy versus toxicity of treatment (Fig. 4-2). The therapeutic index is the “holy grail” of cancer therapy, and it takes careful treatment planning to achieve maximal tumor cell destruction while also sparing normal tissue in hopes of preserving function. There are a host of technologic factors that affect the therapeutic ratio, one in particular is the ability to correctly identify tumor tissue and normal tissue. Use of PET, MRI, and SPECT imaging, as described earlier, allows radiation oncologists to better differentiate target tissue from other tissues. The ability to precisely deliver radiation through techniques such as intensity-modulated radiation therapy (IMRT), stereotactic procedures, and proton therapy permits radiation oncologists to avoid normal tissue while targeting tumor tissue. Increasingly accurate delivery of radiation via image guidance (IGRT) enables smaller margin expansions, thus limiting the size of the dose that reaches normal tissue. However, because of the anatomic location of some tumors, there are still technological limits as to what can be accomplished with radiation alone. Additional improvements are likely to rely on the interaction of systemic agents with technologically advanced radiation delivery methods.

Figure 4-2 Graph of an idealized therapeutic index curve. The therapeutic index is indicated as the difference in probability of tumor response and probability of normal tissue toxicity. The greater the separation of these two curves, the greater the therapeutic index.

General Strategies to Improve the Therapeutic Index

In 1979, Steel and Peckham ⁷ defined four potential ways that combined modality therapy might be used to improve the therapeutic index: (1) independent toxicity, (2) normal tissue protection, (3) spatial cooperation, and (4) enhanced tumor response. As discussed later, the first theoretical concept may not function as originally intended. However, the latter three concepts are relevant for modern strategies of combining drugs with radiation. Recent innovative concepts of biologic cooperation and temporal modulation that expand on Steel and Peckham’s “exploitable mechanisms in combined radiotherapy-chemotherapy”⁷ are having an impact on current investigative strategies for improving the therapeutic index.

Independent Toxicity

To improve the therapeutic index, Steel and Peckham suggested selecting a chemotherapeutic regimen with a toxicity profile distinctly different from that of radiation. This ideal selection of nonoverlapping toxicities could allow for increased tumor cell kill with minimal combined tissue toxicity. Although this has been attempted, success in finding independent toxicity has been elusive. However, the inverse relationship has been implemented to a great extent. It has become a standard of care practice to avoid drugs with toxicities that overlap with radiation; for instance, methotrexate is not used with cranial irradiation, Adriamycin is not used with breast irradiation, and bleomycin is not used with lung irradiation.

Normal Tissue Protection

Only a few clinically relevant, therapeutic agents have been identified as promoting normal tissue protection without protecting tumors. There has been limited success with the free-radical scavenging agent amifostine (WR-2721). The drug appears to be selectively taken up by normal tissue relative to tumor tissue, where it is converted into the active thiol metabolite WR-1065.⁸ Although amifostine has been shown to protect against xerostomia in head and neck cancer treatment (Table 4-1) and to limit renal toxicity from cisplatin, several clinical trials have failed to show an advantage of amifostine use. Undoubtedly, clinical trials will continue to investigate novel radioprotectors that have the potential to affect the therapeutic ratio.

TABLE 4-1 Possible Drug-Radiation Interactions

Mechanism	Example	Notes
Normal tissue protection	Amifostine in head and neck cancer	Reduces xerostomia rates from RT alone
Spatial cooperation	Early-stage breast cancer with adjuvant CT PCI in SCLC	RT provides locoregional control for breast cancer but no impact on DM SCLC CT does not effectively cross BBB → RT can effectively treat the brain
Biologic cooperation	Targeted therapies inhibit prosurvival/proliferation pathways within tumors	Kinase-targeted agents, including tyrosine kinase inhibitors such as dasatinib and sunitinib, as well as monoclonal antibodies such as cetuximab and bevacizumab; mTOR inhibitors
Temporal modulation	Drugs that affect tumor response between fractions, namely, targeting repair, repopulation, reoxygenation, and redistribution	This is essentially a composite of several of the other mechanisms but requires concomitant, rather than sequential, delivery of the drug
Increased DNA damage	Drugs that incorporate into DNA	5-FU and platinum are classic examples
Inhibition of DNA repair	DNA intercalators and nucleoside analogs can disrupt repair and enhance radiation cytotoxicity	Alkylators, antimetabolites, platinum, and topoisomerase inhibitors are a few examples
Cell cycle effects	Most chemotherapeutics are cell cycle specific (except alkylators) Cell cycle arrest in radiosensitive phases (microtubule-targeting agents at M phase) Elimination of radioresistant cells (S phase)	Taxanes, epothilones, 5-FU, gemcitabine, topoisomerase inhibitors are good examples
Targeting repopulation	Conceivably, any systemic agent that has at least cytostatic properties can prevent repopulation	Molecularly targeted agents as well as chemotherapeutics (particularly antimetabolites) can function this way
Hypoxia targeting	Mitomycin C and tirapazamine selectively targeting hypoxic cells Tumor shrinkage by CT decreases interstitial pressure and improves oxygenation	Taxanes and other chemotherapeutics that can produce tumor shrinkage are indirect means (given as induction therapy), whereas mitomycin C and tirapazamine are directly affecting hypoxic cells
Tumor microenvironment targeting	Anti-angiogenesis promotes vascular renormalization	Bevacizumab in glioma

BBB, blood-brain barrier; CT, chemotherapy; DM, distant metastases; 5-FU, 5-fluorouracil; PCI, prophylactic cranial irradiation; RT, radiation therapy; SCLC, small cell lung carcinoma.

Spatial Cooperation

The concept of spatial cooperation implies that chemotherapy and radiation therapy are independent “players,” with chemotherapy acting systemically (i.e., targeting micrometastatic disease) and radiation therapy acting locoregionally. Because these therapies function independently, it could be assumed that a full dose of each would be required to achieve the desired outcome. If the chemotherapy drug and irradiation did function completely independently, then concurrent administration should be possible with nonoverlapping toxicities. It is unclear whether a completely independent action can actually be achieved with the chemotherapeutics currently used with radiation treatment. Some in-field toxicities do occur, suggesting some level of localized radiation sensitization. Therefore, sequential therapy is probably the best means to exploit spatial cooperation. There are many clinical examples supporting this approach, such as adjuvant chemotherapy followed by irradiation for breast cancer, consolidative irradiation for bulky disease in lymphoma, and prophylactic cranial irradiation in small cell lung cancer (see Table 4-1).

Enhanced Tumor Response (Cytotoxic Enhancement)

Current investigative efforts focused on achieving cytotoxic enhancement with combined-modality treatment seek to determine the combination of therapies that leads to an interaction on some level that generates an improved antitumor effect relative to each treatment alone. Some clinical examples have shown that subtherapeutic, or radiosensitizing, doses of chemotherapy can affect distant disease control, suggesting either that increased locoregional control can diminish distant metastatic disease potential or that lower-dose chemotherapy can treat micrometastatic disease (i.e., spatial cooperation).

Biologic Cooperation

The term biologic cooperation is a more recent concept ⁹ that involves independent targeting of subpopulations of cells within the tumor itself (see Table 4-1). Although similar to the concept of spatial cooperation, biologic cooperation implies that some portion of the actual radiation target (i.e., in-field) is resistant to radiation and that this portion becomes the target of the drug given concomitantly. The most prominent example of an agent used in biologic cooperation is the hypoxic cell cytotoxin, tirapazamine. Hypoxia is a known condition of radiation resistance; tirapazamine targets the hypoxic subpopulations of cells because it is most potent in anoxic conditions. Tirapazamine is discussed in more detail later in this chapter.

Temporal Modulation

The four Rs of classical radiobiology, reoxygenation, repair, redistribution, and repopulation,¹⁰ refer to factors that are particularly important for fractionated radiation therapy. For example, antiproliferative therapies could prevent accelerated repopulation between fractions that might not be detectable using single-fraction assays in vitro. Conversely, although the DNA damage repair blockade may enhance radiation sensitivity in a tumor, if DNA repair is also inhibited in normal tissue, outcomes may be worse in fractionated therapy.⁹ Depending on which factors are most prominent in normal and tumor cells, the therapeutic index can be shifted in either a beneficial or a detrimental direction. Therefore temporal modulation implies therapeutics that optimize the four Rs between fractionated radiation treatments⁹ (see Table 4-1).

Potential Biologic Mechanisms of Drug-Radiation Interactions

The many potential mechanisms by which a drug may affect radiation efficacy are summarized in Table 4-1. The classic definition of a radiosensitizer had an enhancement of DNA damage as the critical factor. However, with increased understanding of cancer cell biology, it is apparent that targets other than DNA damage can enhance radiation efficacy. Therefore a more broadly defined “radiation enhancer” can affect several potential mechanisms to increase the effect of radiation.

Increasing Radiation Damage

A drug that enhances DNA damage does so by incorporating itself into the cell’s DNA, thereby increasing the susceptibility of the DNA to radiation damage. Examples of this type of interaction include 5-FU and cisplatin.

Inhibition of DNA Repair

Cancer cells that can effectively repair DNA damage are resistant to the effects of radiation, so compounds that interfere with the DNA damage repair signal transduction pathway can potentially enhance radiation damage. Several chemotherapeutic agents target this process, particularly those that disrupt nucleotide biosynthesis and use. Modified nucleotides include 5-FU, bromodeoxyuridine, gemcitabine, fludarabine, methotrexate, etoposide, and hydroxyurea as well as cisplatin.

Cell Cycle Effects

The G₂/M phase has been identified as the most radiation-sensitive phase of the cell cycle and the S phase as the most radiation-resistant phase.^11,¹² In addition, many cytotoxic chemotherapeutic agents are cell cycle specific. Therefore agents that can maintain cells in radiation-sensitive phases or eliminate those cells in radiation-resistant phases will cooperate with radiation for enhanced efficacy. Although this is clearly seen in preclinical settings, there is much less direct evidence for this phenomenon in clinical data. Nevertheless, taxanes, nucleoside analogs, and modified pyrimidines appear to work in this manner.¹³^–¹⁶

Repopulation

In normal adult tissue, the rate of cell loss is balanced by that of cell proliferation. When increased cell loss occurs from injury, including radiation treatment, signaling for proliferation occurs, resulting in repopulation. Cancers, by their very nature, have an excess of cell proliferation relative to cell loss. When a subtotal cell loss occurs during fractionated radiation, cancers can promote increased proliferation, known as accelerated repopulation. Chemotherapeutic agents with cytotoxic or even cytostatic effects given concurrently with radiation can counteract this repopulation and enhance efficacy.

Hypoxia and Tumor Microenvironment

Solid tumors, particularly those that have grown to any significant size, will contain regions of lower oxygen tension caused by limitations in vascular flow as well as oxygen diffusion within the tumors. Although many tumors trigger angiogenic factors within themselves, these stimulants manifest as aberrant vasculature, often with disorganized architecture. Larger tumors may also have increased interstitial pressure, which leads to a further collapse of blood vessels, creating hypoxic regions and overt necrosis at times.

Hypoxia is one of the most potent factors of radiation protection known. Radiation relies on the production of oxygen free radicals, and hypoxic tissue is two to three times less sensitive to radiation than nonhypoxic tissue.¹⁷ There are four general chemotherapeutic approaches to mitigate hypoxia and enhance radiation efficacy:

1 Chemotherapy shrinks the tumor through cytotoxic action, thereby decreasing interstitial pressure. Because chemotherapy typically targets the fastest-proliferating cells, those cells located next to blood vessels are removed, bringing the hypoxic regions closer to the oxygenated region. A good example of this process is demonstrated by taxanes such as paclitaxel.¹⁸

2 Anti-angiogenic therapies such as bevacizumab—an antibody that targets vascular endothelial growth factor (VEGF)—can potentially normalize vascular flow by eliminating the aberrant neovasculature of the tumor. Work by Rakesh Jain ^19,²⁰ and others^21,²² provides evidence of this phenomenon.

3 Hypoxic cell targeting agents, such as tirapazamine, can provide biologic cooperation by eliminating the most radiation-resistant cells.

4 Hypoxic cell radiation sensitizers can reverse the inherent radiation resistance of hypoxic cells. Drugs such as misonidazole, a nitroimidazole compound, can mimic the effects of oxygen within the hypoxic regions.23,24

Cell Death Pathway Effects

All of the above potential mechanisms of drug-radiation interaction display their efficacy through the consequences of cytotoxicity. However, recently it has become clear that there are several ways in which cytotoxicity is manifest. The Nomenclature Committee on Cell Death (NCCD)’s 2005 classification system,²⁵ based purely on morphologic criteria, defined four modes of cell death: (type 1) apoptosis; (type 2) autophagy; (type 3) necrosis, or oncosis; and (type 4) mitotic catastrophe. Although these are distinct forms of cell death, the stimuli and processes involved are interrelated. There is also evidence that ionizing radiation can manifest its cytotoxicity by each type of cell death. As the understanding of the cell death pathways improves, novel therapeutics targeting each type could enhance radiation efficacy. The cell death mechanisms are briefly described below.

Apoptosis

Apoptosis is the most clearly defined and studied mechanism of cell death. This “programmed cell death” involves characteristic morphologic changes, including chromatin condensation (nuclear pyknosis) and nuclear fragmentation (karyorrhexis). Apoptotic bodies ultimately form, and the cell is removed through phagocytosis without generation of an inflammatory response. Apoptosis can occur with or without caspase activation ^26,²⁷ and does not require DNA fragmentation,²⁵ though this is a classic hallmark. Apoptosis is considered the major mechanism for chemotherapy-induced cell death. As a mechanism for radiation-induced cytotoxicity, apoptosis occurs readily in “liquid tumors” as opposed to most solid tumors, where apoptosis is a minor component of cell death. As such, drugs targeting the apoptosis pathway may enhance radiation cytotoxicity in solid tumors.

Autophagy

Whereas apoptosis is a clear self-destructive mechanism for a cell, the role of autophagy in cell death is more controversial. Autophagy, literally meaning “self-eating,” can provide a protective mechanism for a cell during times of stress (such as nutrient deprivation) because it allows recycling of cellular building blocks through controlled breakdown of cytoplasmic components. However, autophagic cell death does occur; it differs from apoptosis mainly because of the lack of chromatin condensation.²⁵

Necrosis

Type 3 cell death, or necrosis, is a mechanism in which a cell swells (oncosis) until the plasma membrane ruptures and the cell contents are released, resulting in a local inflammatory response.²⁵ The best example of this type of cell death is ischemic injury. Large, single-fraction radiation or radio-ablative doses can produce this type of cell death (e.g., stereotactic radiosurgery of brain lesions).

Mitotic Catastrophe

Mitotic catastrophe is a unique form of cell death that involves failed mitotic events.²⁵ This typically manifests as micronucleation and multinucleation, suggesting that a series of mitotic divisions occur without cytokinesis, which ultimately leads to cell death.

Analyzing Drug-Radiation Interactions

Several methodologies for determining the interaction between a drug and radiation have been detailed in the literature. Several definitions for possible interactions have also been described. The concept of radiosensitization originated many years ago, and classic radiosensitization has been defined as an increased amount of radiation-induced cell death that results from exposure to a second agent, after correction for the cytotoxicity of the agent. Clonogenic survival assay, which measures all forms of cell death as well as prolonged or irreversible cell cycle arrest, is the most encompassing method of measuring radiation cytotoxicity in vitro (Figure 4-3). Survival curves are generated by plating known quantities of cells, treating them with various doses of radiation and/or drug, and plotting the surviving fraction of colonies formed in a semilogarithmic fashion. Normalization is performed by dividing the surviving fraction for treated groups by the plating efficiency, which is defined as the surviving fraction of the untreated cells. Modification in radiosensitization, therefore, is demonstrated in clonogenic survival curve data in which a downward or leftward shift of the normalized surviving fraction implies a radiosensitizing interaction and an upward or rightward shift implies a radioprotective interaction. Although survival curves can show interaction between chemotherapy and irradiation, a better description of radiation modulation is necessary because neither chemotherapy nor radiation cytotoxicity typically follows a linear relationship.

Figure 4-3 Graph of the concept of radiation modulation. The solid line indicates the control clonogenic survival assay plotted as surviving fraction relative to dose in Gy. A combined treatment that causes a rightward shift of the curve indicates a radioprotection effect, whereas a leftward shift of the curve indicates a radiosensitization effect.

One of the early attempts at providing a more descriptive system was by Tyrell.²⁸ His system may be a better starting point for describing various interactions among therapies:

Antagonism: used in all cases in which the action of two treatments that have been combined is less than would be expected from the independent effect of each treatment when it has been used alone.

Zero interaction: used in all cases in which two treatments that have been combined have the effect that would be expected from the independent effect of each treatment when it has been used alone.

Positive interaction: used in all cases in which the action of two treatments that have been combined is greater than would be expected from the independent action of each treatment when it has been used alone.

Synergism: a special case of a positive interaction; the term is used only when kinetic data are available.

These terms seem to have been supplanted by the “additivity” descriptors, including supra-additive, additive, and infra-additive. Once again, the classic paper by Steel and Peckham ⁷ describes the construction of an “envelope of additivity” for evaluating the interaction of two treatments using isobologram analysis. This envelope of additivity is constructed from cytotoxicity data by calculating a mode 1 curve that assumes that both agents have completely independent mechanisms of action as well as a mode 2 curve that assumes that the two agents have exactly the same mechanism of action (see Expert Consult website for isobologram construction detail). When combination therapy data points are plotted on the isobologram, they may fall between mode 1 and mode 2 (additive interaction; within the envelope), above mode 1 (infra-additive interaction), or below mode 2 (supra-additive, or synergistic, interaction). An idealized isobologram is shown in Figure 4-4, and a step-by-step method for constructing an isobologram is included on the Expert Consult website.

Figure 4-4 Graph of an isobologram for examining the interaction of radiation (RT) and a drug. Isoeffective doses of A (RT) and B (Drug) are indicated on the axes. This diagram shows regions of infra-additivity and supra-additivity as well as an envelope of additivity.

The following is a step-by-step procedure for calculating isobolograms using Steel and Peckham’s method ^7,²⁹ (see Fig. 4-WO1).

Web-Only Figure 4-1 Step-by-step construction of an isobologram to define an envelope of additivity for two cancer treatments. Isobologram analysis is used to evaluate the interaction of two treatments, and it requires the construction of an envelope of additivity that is bordered by mode 1 and mode 2 lines. The mode 1 line assumes that the two agents have completely independent mechanisms of action, whereas the mode 2 line assumes that the two agents have the same mechanism of action.

Step 1. The investigator must choose to make the assessments at one level of cytotoxicity (e.g., construct an isobologram that represents the interaction of the agents for a cumulative cytotoxicity of 50%, 10%, or 1%). The example in Figure 4-3 depicts the chosen level of cytotoxicity as horizontal line Z: 1% cytotoxicity (0.01 surviving fraction) in this case.

Step 2. Plots are made of dose-response data for both agents. In Figure 4-WO1, the dose-response data for the two agents are represented by curves A and B.

Step 3. The extreme points of the envelope of additivity are determined. Initially, a separate cartesian graph is created. The y-axis represents the dose of agent B and the x-axis the dose of agent A. The first extreme point of the envelope is placed on the y-axis at the dose of agent B alone that causes a specific level of cytotoxicity, as determined by the dose-response curve of agent B, at the intersection of line Z (Fig. 4-WO1). The second extreme point of the envelope is placed on the x-axis, at the dose of agent A alone that results in the level of cytotoxicity at the intersection of the dose-response curve with line Z (Fig. 4-WO1).

Step 4. The mode 1 line is constructed assuming that the agents function independently. The individual dose-response curves are used for this construction. After exposure to dose X of agent A (XA), a level of cytotoxicity is obtained at a point on the dose-response curve that is above line Z. This level of cytotoxicity is identified as Y. Next, the dose of agent B (XB) that is required to produce cytotoxicity equal to the difference in cytotoxicity at line Z and point Y (identified as C) is determined. The cartesian coordinate (XA, XB) is plotted and becomes a point on the mode 1 line. The entire mode 1 line is constructed in this manner by varying the dose of agent A (for a resulting level of cytotoxicity that falls above line Z) and subsequently calculating the appropriate complementary dose of agent B as described.

Step 5. The mode 2 line is constructed assuming that the two agents have the same mechanism of action. As for mode 1 line construction, exposure to dose X of agent A (XA) results in a level of cytotoxicity identified as Y. The dose-response curve for agent B is then examined. The dose of agent B required for cytotoxicity equivalent to Y is determined and identified as YB. The change in dose of agent B that is required to increase cytotoxicity from YB to line Z is determined and labeled ΔB. The cartesian coordinate (XA, ΔB) is plotted. Similar points are calculated for various initial doses of agent A, and the mode 2 line is formed. The mode 2 line varies in shape depending on whether agent A or agent B is selected first for step 5. Generally, the mode 2 line that results in the greatest separation from the mode 1 line is chosen for the envelope of additivity.

Step 6. The two agents are given concomitantly, and a dose-response curve for concomitant treatment is obtained (typically by holding the dose of one agent constant while varying the dose of the other). The doses of the individual agents that result in combined cytotoxicity equivalent to the level represented by line Z are plotted (J, P).

This procedure allows for characterization of experimental data. The experimental point (J, P) represents an antagonistic interaction if the point falls above the envelope of additivity. The effect of the combination treatment is less than would be expected if the agents had completely independent mechanisms of action. An experimental point that falls directly on the mode 1 line suggests that the agents have independent mechanisms of action and the interaction is additive. An experimental point that falls below the mode 2 line suggests a synergistic interaction between the two agents for the particular concomitant treatment used.

The most difficult result to interpret is an experimental point that falls within the envelope of additivity. In some respects, the envelope of additivity is a misnomer because experimental points that fall in this range display an interaction that is greater than the additive effect that is achieved if the agents function by completely independent mechanisms. The interaction between the agents may be positive if the agents have independent mechanisms of action, or it may be negative if they have identical mechanisms of action.

Although the isobologram analysis is useful, it is somewhat limited because interactions in each analysis are investigated at a single level of cytotoxicity. The investigation of interactions at several levels of cytotoxicity requires the construction of several envelopes of additivity. The ambiguity associated with experimental points that fall within the envelope can be disconcerting and may lead to erroneous conclusions, especially if several levels of cytotoxicity are not investigated. Other mathematical modeling systems have been developed to assess the interaction of agents that cause cytotoxicity. These assessments aim to account for the kinetics of cytotoxicity of the involved agents and to assess multiple levels of cytotoxicity.

Median Dose-Effect Principle

A mathematical modeling system that has gained fairly widespread use for interactions of cytotoxic agents is the median effect principle of Chou and Talalay.29,30,31 This system was derived from Michaelis-Menten equations and basic mass-action law considerations and has been useful for describing competitive enzyme interactions and interactions of cytotoxic agents. The primary relationship of the median effect principle is described by the following equation: f_a/f_u = (D/D_m)^m, in which D is dose, f_a is the fraction affected, f_u is the fraction unaffected, D_m is the dose required to produce the median effect (50%), and m is a Hill-type coefficient used to describe the sigmoid nature of the curve. For first-order Michaelis-Menten kinetics, m = 1.

The following manipulation of this equation can be performed, with surviving fraction (SF) substituted for fraction unaffected in the last step:

The general equation is y = mx + b.

A plot of log[(1/SF) – 1] on the y-axis and log(D) on the x-axis results in a line with a slope of m and a y-intercept of mlog(D_m). The survival curves for the individual agents and for the combination treatment (the individual agents given together in some fashion) can be fitted to the equation for a line by linear regression. If the interaction of two agents is assessed, three lines (i.e., median effect plots) are produced: one for each agent and one for the combination treatment. A graph of the median effect plots for mock individual agents A and B and for the combination of A and B is shown in Figure 4-5. For the combination treatment, D is the sum of the doses of the two agents given concomitantly; it is helpful to perform the experiments with the two agents given together in a fixed ratio of doses (e.g., 1 : 2). By using various total doses (i.e., the sum), with the agents given in the same ratio, it is possible to determine the contribution of the individual agents to the combination treatment in a later calculation.

Figure 4-5 The hypothetical graph (left) demonstrates the median effect principle analysis for agents A and B given alone or in combination. The combination treatments are given in a fixed ratio so that the individual contribution of each agent to the combined effect can be calculated. The combination index (right) is then calculated at various levels of cytotoxicity as measured by surviving fraction (SF). The areas of antagonism, additivity, and synergism are indicated.

This concept can be visualized in Figure 4-5. For instance, in the case of log[(1/SF) – 1] = 0, where SF is surviving fraction, the corresponding log(D), D indicating the sum of the doses of the two agents, can be calculated from the median effect plot for the combination treatment. An example of an actual combination treatment that has been assessed in this manner is radiation followed by a 24-hour exposure to etoposide.²⁹ In this example, a set of experiments was performed with the dose ratio fixed as 32 Gy to 1 µg/mL of etoposide. In this example, the dose D that resulted in log[(1/SF) – 1] equaling a given value was a combination of radiation and etoposide given in the ratio of 32 : 1. The radiation and etoposide components could be discerned, from the median effect plot of the combination treatment, by dividing the resulting dose into the appropriate components based on the ratio of delivery of the two agents.

Definitions used in the median effect principle include the following:

Mutually exclusive: the agents of interest have similar modes of action and do not act independently.

Nonmutually exclusive: the agents of interest have different modes of action or act independently.

Combination index (CI): the derivation of this index is beyond the scope of this chapter. Calculation of CI allows characterization of an interaction as synergistic (CI < 1), antagonistic (CI > 1), or a summation (CI = 1). Chou and Talalay ³¹ provide a full description.

CI can be calculated for any surviving fraction and for mutually exclusive or mutually nonexclusive interactions. For a mutually exclusive interaction,

For a mutually nonexclusive interaction,

in which

(Dx)₁ = D_m[(1/SF) – 1]^1/m, solving the general equation for agent 1 given alone in a dose x.

(Dx)₂ = D_m[(1/SF) – 1]^1/m, solving the general equation for agent 2 given alone in a dose x.

(Dx)_1,2 = D_m[(1/SF) – 1]^1/m1,2, solving the general equation for the agents given in combination for dose x, which represents the sum of the doses of the agents given in a fixed combination.

D₁ = (Dx)_1,2 × (fraction of the mixture that is agent 1).

D₂ = (Dx)_1,2 × (fraction of the mixture that is agent 2).

CI represents the doses of the agents required for a given effect when they are given together, divided by the doses required when the agents are given alone; in this way, CI less than 1 represents a synergistic interaction. A diagram of a CI plot for various levels of surviving fraction is shown in Figure 4-5.

Despite the advent of these robust statistical tools for determining the additivity relationship between treatments, the applicability outside of in vitro models is limited based on time and expense necessary to complete dose-response experiments for both drug and radiation. Therefore preclinical in vivo experimentation typically involves the use of a single drug dose at a concentration that can be achieved clinically.

Chemotherapy and Radiation Therapy and Combinations of Cytotoxic Agents

General Concepts

From the Bench to the Clinic

Occasionally, the process of quantifying interactions of chemotherapy and irradiation has frustrated clinicians attempting to interpret in vitro and in vivo laboratory information, as exemplified by Charles Moertel (quoted by Tannock ³²) in his keynote speech at the first International Conference on Combined-Modality Therapy in 1978:

Based on the results of various individual studies, I could conclude that it is most ideal to administer the nitrosourea 15 hours before irradiation, 2 hours before irradiation, simultaneously with irradiation, or 6 hours after irradiation. While we will continue to cheer our radiation biology colleagues on from the sidelines, I am afraid that we are not yet at the stage where we can comfortably incorporate their results into our clinical practice.

This was not meant as disrespect for the radiobiology community but to point out that, at that time, the laboratory models were potentially quite different from the clinical setting. Because it has been difficult to extrapolate from laboratory results to clinical results, many clinicians have used combination treatments on a trial-and-error basis. However, the reverse order of study has occasionally been fruitful, and efficacious combinations of treatments demonstrated in clinical studies have inspired laboratory investigations that revealed interesting molecular bases of interaction.^29,³⁰ Translational research ideally occurs with a concept that arises from laboratory findings and subsequently is shown to have clinical efficacy. However, preclinical model systems have not always allowed investigators to take findings from the laboratory to the clinic, as indicated by the quotation of Moertel and by many of the early hypoxic cell sensitizer studies.

Therapeutic Benefits

Tannock ³² mentioned another problem with translating findings from the laboratory to the clinical setting, emphasizing that investigators must not merely explore combinations of therapeutic agents to find synergistic interactions but must also find interactions that will produce a therapeutic benefit (e.g., provide greater cytotoxicity in tumor cells than in normal cells). To categorize potentially exploitable differences, Tannock³² described three main categories of biologic diversity between tumor cells and normal cells: tumor cells may display genetic instability compared with normal tissues; tumor cells and normal cells may be different with respect to cellular proliferation or proliferation that occurs after treatment; and environmental factors such as oxygenation and pH may affect tumor cells and normal cells differently. As findings are translated from the laboratory to the clinical setting, it is important to consider the effects of the host mechanisms on these three areas.

Chemotherapeutic Classes

In this section, several classes of systemic agents will be presented, followed by a brief review of clinical data describing combination treatment of these agents with radiation. There are a host of chemotherapeutic classes that are used in patients that will undergo radiation treatment. Although not all of these agents are used concurrently with radiation, it is helpful to understand their predominant mechanisms of action. Below is a brief description of several of the major classes of chemotherapeutics with some information regarding possible means of interaction with radiation. In addition, Figure 4-6 summarizes the cell cycle phase specificity of these agents.

Figure 4-6 Diagram of the cell cycle with the cell cycle phase dependence of various chemotherapeutics.

Antimetabolites

The origin of antimetabolite chemotherapy dates back to the 1940s, when aminopterin was used to treat pediatric leukemia.³³ Since then, a large number of antimetabolite chemotherapeutics have been developed with tremendous success. The targets for these drugs include folate metabolism and nucleoside analogs. The major antimetabolites are presented below.

Fluoropyrimidines: Fluorouracil, Fluorodeoxyuridine, Capecitabine

The fluoropyrimidines, as the name implies, are halogenated pyrimidines that function as antifolates by inhibiting thymidylate synthesis. As mentioned previously in historical perspectives, 5-FU is one of the most established drugs used in combination with radiation. It has been used in both a bolus infusion as well as a continuous venous infusion when combined with radiation and appears to target the radioresistant cells in S phase.¹⁶ The two delivery methods have some differences in terms of side effect profile, but both seem to have good efficacy. In a phase III rectal cancer postoperative adjuvant chemoradiation trial, concurrent continuous infusion of 5-FU during external beam radiation therapy (EBRT) was more effective than the bolus delivery.³⁴ Moreover, data show that 5-FU plasma levels and intracellular metabolite levels are rather short-lived,³⁵ also suggesting a need for continuous administration of the drug for it to be effective with radiation. Because of this, oral formulations have been developed, most notably capecitabine, a fluoropyrimidine carbamate prodrug of 5-FU, which must be converted through the action of thymidine phosphorylase. In addition to the improved patient comfort of taking an oral medication rather than having an infusion pump, another potential advantage of capecitabine in combination with radiation is that it appears that radiation increases thymidine phosphorylase levels in tumors, allowing potential bioaccumulation of active metabolite within irradiated tumor.^36,37

Gemcitabine

Gemcitabine is an analog of deoxycytidine that specifically functions during the S phase by preventing dNTP production. There is both preclinical and clinical evidence demonstrating dramatic radiation-sensitizing properties to combined gemcitabine and radiation.38,39–41 In fact, a significant amount of toxicity has been demonstrated in clinical trials for pancreatic cancer,⁴² necessitating either a decreased dose of gemcitabine or limited field size for radiation treatment ports.

Antifolates: Methotrexate, Trimetrexate, Pemetrexed

The antifolate methotrexate tightly binds dihydrofolate reductase (DHFR), thereby inhibiting folate metabolism. Through this inhibition, thymidylate synthesis is blocked and, therefore, purine biosynthesis as well. In addition, some amino acid synthesis is impaired through blockade of this enzyme, resulting in cytotoxicity.⁴³ Pemetrexed is a pyrrolopyrimidine that functions as an antifolate that inhibits multiple enzymes, including thymidylate synthase, dihydrofolate reductase, glycinamide ribonucleotide formyltransferase, and aminoimidazole carboxamide formyltransferase in a cell cycle–independent manner. Pemetrexed is effective against many solid tumors and has shown radiosensitizing properties in preclinical systems.^44,45,46

Alkylating Agents

Alkylating agents are composed of several classes of electrophilic compounds that share the antitumor characteristic of forming covalent bonds with (“alkylate”) DNA bases. Some alkylators interact with a single strand of DNA whereas others can cross-link two strands. The effect on the DNA leads to the cytotoxicity of these agents. The various classes of alkylators that are potentially used with irradiation are briefly presented here.

Nitrogen Mustard: Chlorambucil, Melphalan

Mustard gas was the first in a long series of alkylating agents that were developed for clinical use. During World War I, people and animals were exposed to mustard gas and were then followed clinically. The effects, particularly those on bone marrow, were noted.33,47 Later, nitrogen mustards were developed for lymphoma therapy as mechlorethamine (Mustargen), eventually used in the MOPP (Mustargen, vincristine [Oncovin], procarbazine, prednisone) regimen for Hodgkin’s lymphoma. The most commonly used nitrogen mustards are chlorambucil and melphalan (L-phenylalanine mustard); they are principally used for the treatment of chronic lymphocytic leukemia and multiple myeloma, respectively. Because of the bone marrow effect of these drugs, caution should be used when irradiating large volumes of bone marrow.

Oxazaphosphorines: Cyclophosphamide, Ifosfamide

The oxazaphosphorines are nitrogen mustard–like compounds that include cyclophosphamide and its structural isomer ifosfamide. These agents are used in combination with radiation in both pediatric and adult cancer treatment.

Mitomycin C

Mitomycin C is an antibiotic with alkylating characteristics derived from Streptomyces. This agent is an aziridine ring–containing compound that resembles nitrogen mustard. Mitomycin C blocks DNA synthesis but also causes cell cycle arrest during G₂/M-phase transition.^48,⁴⁹ Mitomycin C has also been shown to function well as a hypoxic cell radiosensitizer,⁵⁰ which may help to explain why mitomycin C–based chemoradiation is so effective in anal cancer treatment,^51,⁵² as discussed later in the chapter.

Triazenes: Procarbazine, Dacarbazine, Temozolomide

Temozolomide has revolutionized the treatment of high-grade gliomas when it is used in combination with irradiation. It effectively crosses the blood-brain barrier (cerebrospinal fluid levels may be 30% of plasma levels).⁵³ Temozolomide generates DNA damage through methylation of DNA at the O-6 position of guanine. Interestingly, O-6-methylguanine-DNA methyltransferase (MGMT) is a p53 DNA repair enzyme that, when its gene is methylated (i.e., epigenetically silenced), appears to predict for response to temozolomide-based chemoradiation because of the inability of the cancer cell to remove the O-6-methylguanine causing cytotoxicity.⁵⁴ The DNA methylation from temozolomide triggers the mismatch repair pathway with a G₂/M-phase arrest, yielding apoptosis and radiosensitive cells.⁵⁵ Ongoing clinical trials are examining the importance of MGMT status and the combination of temozolomide with other agents.⁵⁴

Nitrosoureas: BCNU (Carmustine), Methyl CCNU (Semustine), CCNU (Lomustine), Streptozocin

Several members of the nitrosourea group of alkylating agents are capable of crossing the blood-brain barrier and cross-linking DNA. BCNU (carmustine), has been used to treat brain tumors, predominantly gliomas, but has also been used for multiple myeloma and in high-dosage transplant regimens. Gliadel, a BCNU-impregnated wafer, can be placed in a glioma resection cavity, where it biodegrades slowly to release the chemotherapeutic agent. CCNU (lomustine) is a related compound with increased lipid solubility and is also used for brain tumors.⁵⁶

Platinum Agents: Cisplatin, Carboplatin, Oxaliplatin, Satraplatin

Cisplatin (cis-diamminedichloroplatinum[II]) is the prototypical, most widely studied member of the platinum family. It has been used for decades as an anticancer treatment. Preclinical work in the late 1970s by Soloway and colleagues ⁵⁷ demonstrated radiosensitization in a murine model of transitional cell carcinoma. Since then, both clinical and preclinical data have indicated several mechanisms of interaction between cisplatin and radiation. Potential cooperation between the two modalities can occur at the level of DNA because radiation often causes repairable single-strand breaks in DNA, which can be converted to lethal double-strand breaks when they occur close to cisplatin-DNA adducts (intrastrand and interstrand cross-links). Cisplatin’s function as a free-electron scavenger may also impair the DNA repair mechanism and, therefore, the ability “fix” the radiation-induced DNA damage.⁵⁸ In addition, radiation may enhance the uptake of cisplatin into the cell as well as help generate active platinum metabolites.¹ The other family members—carboplatin, oxaliplatin, and the orally active agent satraplatin—appear to have similar mechanisms of action. Differences among the members may be the result of the three-dimensional structure of the DNA adducts that each platinum agent generates, which influences binding to various polymerases and DNA repair enzymes.⁵⁹ For these reasons, the platinums function independently of the cell cycle phase.

Taxanes: Paclitaxel, Docetaxel, Albumin-Bound Paclitaxel

Unlike the vinca alkaloids, the taxanes stabilize microtubules and promote further tubulin polymerization, which inhibits centrosome mechanics during mitosis. In terms of enhancing radiation effect, the taxanes appear to manipulate several of the factors listed in Table 4-1. First, the taxanes can block the metaphase-anaphase cell cycle checkpoint, which could allow for accumulation of cells in the radiosensitive G₂/M phase.^61,62 Furthermore, taxanes can cause tumor shrinkage,^63,⁶⁴ thereby decreasing interstitial pressure and allowing improved oxygenation.¹⁸ In addition, taxanes can manipulate signal transduction cascades involved in radiation response.⁶⁵

Epothilones: Epothilone B, Aza-epothilone B (Ixabepilone)

The epothilones are considered next-generation microtubule-targeting agents with functions similar to that of taxanes but derived from Mycobacterium. They stabilize microtubules with high potency and halt mitosis in a manner similar to taxanes.⁶¹ These drugs were developed to be independent of the p-glycoprotein efflux–resistance mechanisms that target taxanes and vinca alkaloids.^61,66 These drugs are likely to become a popular choice for concurrent chemoradiation regimens in the future.

Topoisomerase Inhibitors

Topoisomerases are critical enzymes in DNA replication of all cells because of their ability to unwind DNA. There are two major classes of topoisomerases in mammalian cells that are clinically relevant for oncology therapeutics: topoisomerase I and II. Topoisomerase I (TopI) is involved in DNA replication fork movement and unwinding supercoils during DNA transcription, whereas topoisomerase II (TopII) is important for untangling DNA during transcription and remodeling of chromatin.^35,⁴³ These classes are named based on how many DNA strand breaks are created during enzymatic action; that is, a single-strand break for TopI and a double-strand break for TopII.⁴³ These breaks are required for TopI and TopII to unwind and disentangle DNA but are temporary, because the enzyme will reconnect the broken strands (re-ligation). TopI/II inhibitors cause cytotoxicity by disrupting the process and generating DNA double-strand breaks.

Topoisomerase I Inhibitors

Camptothecins Irinotecan, Topotecan

Camptothecin is a naturally occurring alkaloid derived from the plant Camptotheca acuminata that was identified in an anticancer drug discovery screen in the 1960s.⁶⁷ Camptothecin forms a stable ternary complex, preventing normal DNA re-ligation, and causes the complex to collide with the replication fork, leading to a DNA double-strand break and cytotoxicity.⁶⁸ The S-phase specificity of this drug class provides some of the rationale for radiosensitization. The drug was unsuccessful in the clinic because it caused severe urinary complications, though camptothecin is still used as a research tool and for positive control for cytotoxicity and apoptosis. However, derivatives of camptothecin, notably irinotecan and topotecan, are used as chemotherapeutic agents. Irinotecan is approved by the Food and Drug Administration (FDA) to treat colorectal cancer, but data related to concomitant administration with radiation have also been generated in both small cell⁶⁹^–⁷¹ and non–small cell lung cancer patients.^72,⁷³ Topotecan is approved for ovarian, small cell, and cervical cancer. It has been combined with radiation to treat glioblastoma in clinical trials.^74,⁷⁵

Topoisomerase II Inhibitors

Podophyllotoxins: Etoposide, Etoposide Phosphate, Teniposide

The plant extract podophyllotoxin has microtubule-binding activity. The clinically used derivatives do not function through microtubule action but actually are TopII poisons.⁷⁶ These epipodophyllotoxins, most notably etoposide and teniposide, are glycoside derivatives used to treat both childhood and adult tumors.³⁵ In addition, these drugs have been used with radiation in both sequential and concomitant regimens.^77–79,80

Anthracyclines: Idarubicin, Doxorubicin, Epirubicin, Daunorubicin

Anthracyclines are naturally occurring substances that intercalate into the DNA when they target TopII, leading to DNA double-strand breaks.^35,⁸¹ These drugs have a wide range of clinical indications, including both liquid and solid tumors. In terms of radiation sensitization, the interaction of doxorubicin and radiation is well known, so concurrent administration is generally avoided. In fact, when doxorubicin is given after radiation, an inflammatory reaction known as radiation recall can occur.^82,⁸³

Other Agents: Mitoxantrone, Dactinomycin

Mitoxantrone is an anthracenedione designed to function like an anthracycline, but it has less cardiotoxicity ⁸⁴ because it is less likely to form free radicals³⁵ and it may affect calcium release differently.⁸⁵ Like anthracyclines, mitoxantrone can intercalate in DNA and poison TopII to form DNA double-strand breaks. This drug is approved to treat hormone-refractory prostate cancer. Dactinomycin is a Streptomyces-derived antibiotic that can intercalate in DNA, block TopII, and cause DNA double-strand breaks.³⁵ Dactinomycin is used in pediatric sarcoma therapeutic regimens, including treatment of rhabdomyosarcoma.

Chemoradiation for Specific Types of Cancer

An all-encompassing review of the clinical examples of the use of chemoradiation is beyond the scope of this chapter. More comprehensive information regarding disease site–specific trials is included in the individual chapters devoted to each site. What follows is merely a brief account of some of the landmark trials that have demonstrated improved organ preservation, local control, and overall survival rates in the modern era. Specifically, aerodigestive, genitourinary, gynecologic, and central nervous system cancer examples are presented.

Gastrointestinal Cancers

Anal Cancer

A classic example of the evolution of an efficacious interaction of chemotherapy and radiotherapy is combined-modality treatment for anal cancer. In the early 1970s, researchers discovered that anal cancer could be treated successfully with a combination of 5-fluorouracil (5-FU), mitomycin C, and irradiation.⁵² In 1974, Nigro and others⁵² reported on three patients who received variations of these three treatments, with excellent responses to the preoperative therapy (Table 4-2). This article became a classic in the oncology literature, and the regimen became prominent in the treatment of anal cancer. Because of this regimen, many patients were spared abdominal perineal resection.

TABLE 4-2 Combined-Modality Treatment for Anal Cancer: A Study of Three Patients

After the initial report of Nigro and colleagues,⁵² several other groups confirmed the efficacy of chemotherapy and irradiation (without surgery) as the standard treatment for primary anal cancer^52,86,87,88 (Table 4-3). Subsequently, an intergroup effort was undertaken by the Radiation Therapy Oncology Group (RTOG) and Eastern Cooperative Oncology Group (ECOG) to determine whether mitomycin C could be removed from the regimen, because its inclusion resulted in increased toxicity compared with that of 5-FU and radiation without mitomycin C. With 5-FU alone, however, fewer patients were able to avoid colostomy (Table 4-3).

TABLE 4-3 Concomitant Radiation Therapy and Chemotherapy for Anal Cancer

Additional attempts at replacing mitomycin C with less toxic concurrent chemotherapy agents have been undertaken, most notably in the United States. The GI Intergroup study (Table 4-3), coordinated by the RTOG (RTOG 98-11), compared an induction chemotherapy regimen (5-FU and cisplatin) followed by the same chemotherapy concurrently with radiation versus the standard regimen of concurrent chemotherapy (5-FU and mitomycin C) with radiation. The hypothesis was that the induction chemotherapy would decrease the tumor bulk, making radiotherapy more effective and improving local control, and that the additional cycles of induction chemotherapy might improve overall survival rates by decreasing distant metastases. However, these hypotheses were disproven because the cisplatin arm not only failed to show a benefit in terms of local control, disease-free survival rates, and overall survival rates but was clearly inferior to mitomycin C in terms of colostomy-free survival rates.⁵¹ Therefore the combination of 5-FU, mitomycin C, and irradiation remains the standard regimen for anal cancer. A subsequent phase II RTOG chemoradiation trial for anal cancer found that intensity-modulated radiation therapy (IMRT) techniques can reduce the toxicity of the standard of care by limiting normal tissue toxicity.

A discussion of how the laboratory investigations followed the clinical data can be found in the expanded version of this chapter available at the Expert Consult website.

Laboratory Investigations in Anal Cancer

The clinical finding of the efficacious combination of 5-FU, mitomycin C, and radiation led to laboratory studies. Dobrowsky and associates⁸⁹ performed a complex isobologram analysis using the same agents reported by Nigro and colleagues.⁵² The assessment by Dobrowsky and co-workers,⁸⁹ using an in vitro system of a squamous tumor cell line, illustrated some of the difficulties with the ideal progression of taking laboratory discoveries to the clinic. Two different endpoints were used: colony formation (cells plated after treatment and allowed to form colonies) and viable cells per flask (obtained by multiplying the cell number per flask at 96 hours by the surviving fraction, as stipulated by a standard colony formation assay). In an attempt to duplicate the clinical treatment of Nigro and colleagues,⁵² mitomycin C was given as a 1-hour exposure and 5-FU as a 4-day exposure after initial radiation.

The first experiments assessed the interaction of 5-FU and mitomycin C without radiation. Initially, a single dose of mitomycin C (0.5 µg/mL for 1 hour) was combined with various doses of 5-FU. Isobolograms were constructed for the colony formation endpoint at a surviving fraction of 0.04. Isobologram construction showed that the combination treatment resulted in an experimental point below the envelope of additivity at this level of cytotoxic assessment. Isobolograms also were constructed for the viable cells per flask endpoint at the 1% viability level; the experimental point for combined 5-FU and mitomycin C was directly on the mode 2 line. This endpoint was included because it was believed to account for the cytotoxic and cytostatic effects of the treatment.

Because the results of this synergy analysis varied with the endpoint used, the optimal endpoint, whether colony formation or viable cells per flask, is not known. That the use of these slightly different endpoints produced slightly different isobologram results illustrates some of the problems in interpreting in vitro data and in attempting to extrapolate this information to the clinical setting. In the future, it may be possible to assess which endpoints may be most useful for various cytotoxic agents and various tumors on the basis of the relative contribution of cytotoxic and cytostatic effects for a given situation.

On the basis of the experiments without irradiation, specific concentrations of mitomycin C (0.5 µg/mL) and 5-FU (0.15 µg/mL) were selected for subsequent experiments involving radiation;⁸⁹ these concentrations resulted in 60% and 80% surviving fractions, respectively. With colony formation as the endpoint, it was discovered that the interaction of irradiation and 5-FU or irradiation and mitomycin C (at the levels of cytotoxicity assessed) produced experimental points below the envelope of additivity. These results corroborated those reported previously by Byfield and colleagues,⁹⁰ in which some level of 5-FU cytotoxicity was required for a positive interaction with irradiation. However, the results of irradiation in conjunction with mitomycin C were not entirely consistent with those of previous reports, which had suggested that a positive interaction of these agents did not exist.⁹¹

The previous example illustrates several important points. If the protocol of Nigro and associates ⁵² had been designed on the basis of laboratory studies (if all of the just-mentioned studies had existed in 1974), it would have been difficult to assess where to begin. First, the investigator would need to decide which in vitro endpoint would be most relevant to anal cancer (viable cells per flask or colony formation), and this decision would affect whether one believed that 5-FU and mitomycin C interacted synergistically. Second, the investigator would need to decide which assessment of mitomycin C and radiation was most relevant to the treatment of anal cancer, because authors disagreed about whether this interaction was synergistic. This example also illustrates the challenges that are faced when interpretations of in vitro or in vivo experimental data are used to guide the design of clinical trials. These challenges can be exciting as we learn more about the significance of various endpoints at the molecular level and how these molecular events may be manipulated in a particular tumor.

Esophageal Cancer

Esophageal cancer remains a very challenging cancer to treat, primarily because of the locally advanced stage that is typically found at diagnosis. For nonsurgical approaches to treatment, radiation alone has been shown to be quite limited in terms of controlling the disease. Indeed, chemoradiation has been clearly shown to be the treatment of choice following several landmark trials comparing chemoradiation with radiation alone. The Intergroup trial coordinated by RTOG (85-01), originally published by Herskovic and colleagues ⁹² and later updated by Cooper and associates,⁹³ randomized patients to either 64 Gy radiation alone at 2 Gy/fraction or 50 Gy at 2 Gy/fraction concurrent with 5-FU (1000 mg/m²/day for days 1 to 4) and cisplatin (75 mg/m² on day 1). In the concurrent arm, the chemotherapy was given every 28 days during radiation (weeks 1 to 5 of EBRT) and every 21 days thereafter for two additional cycles. This trial established that radiation alone was inferior to combined-modality chemoradiation (5-year overall survival [OS] was 0% vs. 26%; p = .0001).⁹³ A meta-analysis by Wong and associates⁹⁴ confirmed that concurrent chemoradiation was beneficial in terms of survival, with a hazard ratio (HR) of 0.73, 95% confidence interval of 0.64 to 0.84, p <.0001. Sequential chemotherapy and radiation did not show a statistically significant benefit, however.⁹⁴

Recently, Tepper and colleagues ⁹⁵ published the results of the U.S. GI Intergroup study (CALGB 9781) that randomized patients to neoadjuvant cisplatin/5-FU/EBRT (50.4 Gy) before esophagectomy versus esophagectomy alone. Although the trial was closed early because of poor accrual, the 56 patients enrolled were analyzed on an intent-to-treat analysis. This revealed a significant difference in median and 5-year overall survival (OS) with trimodality treatment versus surgery alone (median: 54 mo vs. 21.6 mo; 5-yr OS: 39% vs. 16%, p = .002).⁹⁵ In the future, more trials will combine molecularly targeted agents with standard chemotherapies concurrently with radiation.⁹⁶

Gastric Cancer

The U.S. GI Intergroup 0116 phase III trial, reported by Macdonald and colleagues,⁹⁷ compared adjuvant postoperative chemoradiation with surgery alone for patients with resected high-risk gastric or gastroesophageal cancers. Bolus 5-FU/leucovorin was given before EBRT (one 5-day cycle), concurrently with EBRT (two cycles: 4-day for week 1, 3-day for week 5), and after EBRT (two additional 5-day cycles). A survival advantage of concurrent chemoradiation was shown (3-year OS, 50% vs. 41%, p = .005; 3-year relapse-free survival, 48% vs. 31%, p = .001),⁹⁷ and this treatment became the standard of care for gastric cancer in the United States. In the United Kingdom, the MAGIC trial established a nonirradiation regimen that involves perioperative (neoadjuvant and adjuvant) epirubicin, cisplatin, and 5-FU (ECF) chemotherapy as an appropriate standard of care for resectable gastric cancer.⁹⁸ The recently completed postoperative adjuvant U.S. GI Intergroup phase III trial essentially “married” the INT-0116 trial and the MAGIC trial by investigating the role of chemoradiation in the setting of more modern ECF chemotherapy as the experimental arm versus the control arm of GI INT 0116, and results are awaited.

Rectal Cancer

Although resectable rectal cancer is a surgically managed disease, the addition of adjuvant preoperative or postoperative chemoradiation for high-risk patients is well recognized as a vital component of therapy. Four randomized trials 99,100–102 investigating the addition of chemotherapy to neoadjuvant EBRT in stage II and III rectal cancer are summarized in Table 4-4 (the Bujko trial did not use the same dose/duration of EBRT, so it is not a true comparison of EBRT alone and EBRT with concurrent chemotherapy). Furthermore, a recently published meta-analysis analyzed these four trials.¹⁰³ Although this analysis showed an improved complete pathologic response rate and local control with the addition of chemotherapy to preoperative radiation, no benefit was found in terms of sphincter preservation, disease-free survival rates, or overall survival rates. Of note, when preoperative chemoradiation was given, grades 3 and 4 toxicity were increased over that of preoperative EBRT given alone.

TABLE 4-4 Randomized Trials of Neoadjuvant Concomitant Radiation Therapy and Chemotherapy for Rectal Cancer

The current standard-of-care approach, however, was defined in the phase III German Rectal Trial.¹⁰⁴ Preoperative chemoradiation was shown to be superior to postoperative chemoradiation in terms of local control, sphincter preservation rates, and toxicity.

As in other disease sites, more recent trials are investigating the addition of molecularly targeted agents to the standard preoperative chemoradiation regimen. The antivascular endothelial growth factor antibody bevacizumab has shown promising results in a recently published phase I/II trial in which all 32 patients showed tumor regression following neoadjuvant therapy, with an actuarial 5-year disease-free survival rate of 75%.¹⁰⁵

Head and Neck Cancers

Treatment of locally advanced head and neck tumors was traditionally managed with surgery and postoperative radiation. However, over the past two decades, an explosion of chemoradiation trials shifted management toward an organ preservation approach (summarized in Table 4-5). One of the most impressive results for a randomized trial in head and neck cancer was that of Intergroup 0099 (RTOG 8817), originally published by Al-Sarraf and colleagues in 1998.¹⁰⁶ In this trial of nasopharyngeal cancer patients, radiation alone (70 Gy at 2 Gy/fx) versus radiation with concurrent cisplatin plus adjuvant cisplatin/5-FU demonstrated a dramatic 67% versus 37% 5-year overall survival (OS) advantage in favor of the chemoradiation arm. More than 90 randomized clinical trials have been performed examining chemoradiation in head and neck cancer. Several meta-analyses have been published showing an absolute survival benefit for chemoradiation. Indeed, the most recent update of the MACH-NC in 2009 analyzed over 17,000 patients in 93 randomized trials and showed an absolute overall survival benefit of 6.5% at 5 years.¹⁰⁷ Although two large, randomized trials of neoadjuvant chemotherapy followed by radiation have shown a benefit in terms of laryngeal organ preservation,^108,109 subsequent studies, including RTOG 9111¹¹⁰ and the MACH-NC,¹⁰⁷ suggest that concurrent chemoradiation is more effective than sequential administration. Currently, one of the questions being asked in clinical trials is whether or not induction chemotherapy followed by concurrent chemoradiation will be superior to concurrent chemoradiation alone in locally advanced head and neck cancer.

TABLE 4-5 Selected Concomitant Radiation Therapy and Chemotherapy for Head and Neck Cancer

As mentioned above, the RTOG 9111 was a landmark Intergroup trial for patients with what would currently be staged as T2 and T3 glottic and supraglottic squamous cell carcinomas.¹¹⁰ This trial randomized patients to one of three arms: (1) neoadjuvant chemotherapy, (2) concurrent chemoradiation, and (3) radiation alone. Although overall survival was not different among the groups, the disease-free survival and locoregional control rates favored the concurrent chemoradiation arm.¹¹⁰ This benefit is not restricted to organ preservation studies. Two trials published in the same issue of the New England Journal of Medicine detailed the RTOG¹¹¹ and EORTC¹¹² trials randomizing patients to postoperative radiation with or without concurrent chemotherapy. Although the two trials had slight differences in their inclusion criteria, they both established the importance of adjuvant chemoradiation in the setting of high-risk postoperative patients. A pooled data analysis from the two trials identified positive margins and extracapsular extension as the two significant risk factors for combining chemotherapy with radiation in the adjuvant setting.¹¹³

In 2004, Bonner and associates ¹¹⁴ published the results of a randomized phase III trial incorporating a molecularly targeted therapy with radiation in locally advanced head and neck cancer. This trial used cetuximab, a monoclonal chimeric (mouse and human) antibody to EGFR. This trial showed not only a local control benefit but also an overall survival benefit when cetuximab was combined with radiation, including altered fractionation schedules. In addition, the incidence of serious toxicity (other than rash and transfusion reactions) was similar between the groups; this finding is in stark contrast to chemoradiation regimens, which invariably have increased toxicity. These results have held up in the most recently published update,¹¹⁵ which shows a 49-month median survival rate in the cetuximab arm and a 29.3-month survival rate in the radiation-alone arm. Five-year overall survival rates were 45.6% in the cetuximab group and 36.4% in the radiation-alone group (hazard ratio of 0.73; p = .018).

Several trials are accruing or have recently closed in which cetuximab was combined with chemoradiation, including the phase III RTOG 0522 trial, randomizing patients with stage III and IV squamous cell carcinoma of the oropharynx, hypopharynx, and larynx to cisplatin (100 mg/m² every 3 weeks) with or without a 1-week pretreatment (400 mg/m²) and concurrent (250 mg/m²/week) administration of cetuximab with accelerated fractionation. Those results should be forthcoming.

Non–Small Cell Lung Carcinoma

Non–small cell lung carcinoma (NSCLC) is the number one cause of cancer deaths in the United States. This is at least partly because of the late presentation that typically occurs with this disease (usually, stage III or IV disease is seen).

The treatment of stage III NSCLC has evolved over the past two decades. For unresectable stage III patients, EBRT alone produced poor 2-year overall survival rates on the order of 20%. A modest improvement (up to 29%) was achieved with more intense radiation schedules such as the continuous hyperfractionated accelerated radiotherapy (CHART) regimen.¹¹⁶

To improve upon these outcomes, the approach most heavily tested was the addition of systemic agents to EBRT, which is summarized in Table 4-6. Three major trials of sequential chemotherapy followed by irradiation were published in the 1990s that showed improvement with the addition of platinum-based chemotherapy.^117,118,119 Around the same time, a meta-analysis demonstrated a small but significant improvement with the addition of chemotherapy.¹²⁰ However, because the various trials included many different sequences of chemotherapy and radiation, it was unclear what would be the best regimen. Therefore several randomized phase III trials compared sequential with concurrent chemoradiation.^{121,122,123,124} In all but one trial,¹²⁴ the concurrent arm showed a significant improvement relative to sequential chemotherapy and irradiation. This established concurrent chemoradiation as the treatment of choice. What was not clear, however, was whether adjuvant chemotherapy could be added to concurrent chemoradiation regimens and, if so, whether it would be better to administer the chemotherapy neoadjuvantly or adjuvantly.

TABLE 4-6 Concomitant Radiation Therapy and Chemotherapy for Non–Small Cell Lung Cancer

Study	Regimen	Outcome
Dillman et al¹¹⁷ (CALGB 8433 updated¹⁶⁶)	Sequential cisplatin (100 mg/m² on days 1 and 29) and vinblastine (5 mg/m² weekly × 5) followed by RT (60 Gy) vs. RT alone (60 Gy)	Improved MS (13.7 mo vs. 9.6 mo; p = .012) with CT/RT
Sause et al¹¹⁸ (RTOG 8808 updated¹⁶⁷)	Three-arm trial: sequential cisplatin (100 mg/m² on days 1 and 29) and vinblastine (5 mg/m² weekly × 5) followed by RT (60 Gy) vs. HyperFx RT (69.6 Gy at 1.2 Gy BID) vs. RT alone (60 Gy)	Improved MS (13.2 mo vs. 12 mo vs. 11.4 mo; p = .04) with CT/RT
Furuse et al¹²²	CT of cisplatin (80 mg/m² on days 1 and 29), vindesine (3 mg/m² on days 1, 8, 29, and 36), and mitomycin C (8 mg/m² on days 1 and 29) given either concurrently with split-course RT (28 Gy × 2, 10 days apart) or sequentially with RT (56 Gy)	Improved MS (16.5 mo vs. 13.3 mo; p = .03998) with concurrent CT/RT
Curran et al¹²¹ (RTOG 9410)	Three-arm trial: cisplatin (100 mg/m²) and vinblastine (5 mg/m²) given before (arm 1) or concurrently with once-daily RT (arm 2, 60 Gy), or concurrent cisplatin (50 mg/m²) and oral etoposide (50 mg BID) with HyperFx RT (arm 3, 69.6 Gy at 1.2 Gy BID)	Improved MS (17 mo) in concurrent once-daily CT/RT arm (arm 2) vs. 15.6 mo (arm 3) vs. 14.6 mo (arm 1) (p = .038)
Albain et al¹²⁶ (INT 0139)	Stage IIIA patients received two cycles of CT: cisplatin (50 mg/m² on days 1, 8, 29, and 36) and etoposide (50 mg/m² on days 1-5 and 29-33) plus 45 Gy. If no progression, they were randomized to surgery or 16-Gy boost. All patients received two adjuvant cycles of cisplatin/etoposide.	No difference in MS or 5-yr OS; improved median PFS in trimodality arm (12.8 mo vs. 10.5 mo; p = .017); in subset analysis, improved OS in patients who underwent lobectomy vs. CT/RT, but CT/RT was better if patient underwent pneumonectomy.
Thomas et al¹²⁷ (GLCCG)	Patients received three cycles of cisplatin and etoposide and were randomized to CT/RT with concurrent carboplatin and vindesine followed by S or S followed by RT alone	Preoperative CT/RT resulted in increased pathologic response but no improvement in PFS. For patients requiring pneumonectomy, preoperative CT/RT had trend to increased Tx-related mortality (14% vs. 6%; p = .14).

BID, twice daily; CALGB, Cancer and Leukemia Group B; CT, chemotherapy; CT/RT, chemoradiation therapy; GLCCG, German Lung Cancer Cooperative Group; HyperFx, hyperfractionation; INT, Intergroup; MS, median survival; OS, overall survival; PFS, progression-free survival; RT, radiation therapy; RTOG, Radiation Therapy Oncology Group; S, surgery; Tx, treatment.

A phase II trial called the LAMP (Locally Advanced Multimodality Protocol) trial attempted to determine what would be the best approach.¹²⁵ The three arms in this trial included (1) neoadjuvant chemotherapy followed by irradiation, (2) neoadjuvant chemotherapy followed by concurrent chemoradiation, and (3) concurrent chemoradiation followed by adjuvant chemotherapy. Although the concurrent chemoradiation followed by adjuvant chemotherapy arm had the best outcome, the trial had some limitations, including insufficient power to determine the best regimen.¹²⁵

Although concurrent chemoradiation is the standard approach for stage III NSCLC, there are certain instances in which surgery can be combined with chemoradiation in a trimodality approach. Two major randomized trials have been published regarding this trimodality approach; they are also included in Table 4-6 (U.S. Lung Intergroup 0139¹²⁶ and the German Lung Cancer Cooperative Group¹²⁷ trials). Both trials appeared to show that trimodality therapy was feasible but that preoperative chemoradiation should be avoided if pneumonectomy would be performed because excessive treatment-related deaths occurred following pneumonectomy.

Cervical Cancer

One of the clearest clinical examples demonstrating improved outcomes of combined chemoradiation is in locally advanced cervical cancer. Cisplatin-based chemoradiation regimens have had clear success in a series of randomized phase III trials showing not only improvement in locoregional control but also in overall survival rates. This series of clinical trials (summarized in Table 4-7) had differing inclusion criteria and various treatment approaches, yet these trials presented convincing evidence that cisplatin, when combined with irradiation, improves outcome in locally advanced cervical cancer. An excellent review of the historical context and remaining controversies regarding these trials is available,¹²⁸ so they will be discussed only briefly here.

TABLE 4-7 Concomitant Radiation Therapy and Chemotherapy for Cervical Cancer

The Gynecological Oncology Group (GOG) had three positive trials for chemoradiation as a component of treatment. The GOG 85 trial investigated two different concurrent chemotherapy regimens (hydroxyurea vs. cisplatin/5-FU) with irradiation for patients with locally advanced cervical cancer.¹²⁹ This trial demonstrated that concurrent cisplatin/5-FU was superior to hydroxyurea (5-year OS, 62% vs. 50%, p = .018). GOG 120 was a three-arm trial that compared radiation plus hydroxyurea with radiation plus two different concurrent cisplatin-containing regimens (one with cisplatin alone and the other with cisplatin plus 5-FU and hydroxyurea).¹³⁰ The two cisplatin-containing arms were superior to the control arm (3-year OS for both cisplatin-based arms of 65% vs. 47%, p <.005). Because cisplatin alone was less toxic than the 5-FU/hydroxyurea combination, the concurrent cisplatin/irradiation approach was preferred. The GOG also investigated the importance of chemotherapy with irradiation for bulky stage IB patients treated with hysterectomy in GOG 123. In this trial, the patients were randomized to radiation alone versus radiation plus weekly cisplatin before undergoing hysterectomy.¹³¹ Once again, the cisplatin-containing arm was superior (3-year OS, 83% vs. 74%; p = .008).

In the 1990s, the RTOG also examined the importance of cisplatin with radiation for patients with locally advanced cervical cancer. RTOG 9001 compared extended field radiation alone with pelvic radiation plus cisplatin and 5-FU.¹³² The chemoradiation arm was superior to the extended field irradiation (5-year OS, 73% vs. 58%; p = .004).

The fifth trial to support concurrent cisplatin/irradiation was the Intergroup 0107/SWOG 8797, which tested chemoradiation in the adjuvant setting for patients who underwent hysterectomy with high-risk features at the time of surgery, including positive pelvic nodes, positive margins, and parametrial involvement.¹³³ Patients were randomly assigned to receive radiation alone or radiation with cisplatin/5-FU. The chemoradiation group was again superior (4-year OS, 81% vs. 71%; p = .007).

Only one major trial, from the National Cancer Institute of Canada (NCIC), failed to demonstrate an overall survival benefit for this approach.¹³⁴ However, that trial has been criticized because of the small size and wide confidence intervals that may have prevented a difference from being detected.

In the next generation of chemoradiation trials, molecularly targeted agents, including cetuximab and tirapazamine, are being added to standard cisplatin-based regimens.

Genitourinary Cancer

Chemoradiation is established as the standard bladder-preserving management strategy for muscle invasive bladder cancer. Shipley and associates ¹³⁵ published a single institution experience using a neoadjuvant chemoradiation approach similar to the larynx preservation trials discussed above in which a cisplatin-containing chemoradiation regimen is given (following a maximal TURBT, or transurethral resection of bladder tumor). An evaluation for tumor response is then performed, followed either by surgery (for anything less than a complete response) or by consolidative chemoradiation. With this approach, the 5-year overall survival rate is comparable to that of the surgical series at 54%. Bladder preservation rates without invasive local recurrence are around 50%.

Another genitourinary example involves the use of hormone ablation therapy plus EBRT for high-risk prostate cancer demonstrated to have benefits in local control, disease-free survival rates, biochemical disease-free survival rates, and, possibly, overall survival as opposed to EBRT alone, based on both the RTOG ¹³⁶^–¹³⁸ and EORTC^139,140 trials. It has been postulated that androgen deprivation therapy used in concomitant radiation regimens not only provides a spatial cooperation interaction but also a biologic cooperation function.

Glioblastoma

The treatment of glioblastoma has dramatically changed in recent years based on the encouraging findings from the phase III trial from EORTC and NCIC. This trial, first published by Stupp and colleagues in 2005 ¹⁴¹ and recently updated with 5-year data,¹⁴² demonstrated a remarkable improvement in median survival (14.6 months vs. 12.1 months) and 5-year OS (9.8% vs. 1.9%; HR, 0.63; p <.0001) with the use of concurrent temozolomide and radiation with adjuvant temozolomide. Prognostic^143,144,145 and possibly predictive¹⁴⁶ information can be garnered from evaluating the MGMT methylation status, because patients whose tumors have MGMT epigenetic silencing as a result of hypermethylation do better. Currently, there are several trials evaluating temozolomide, including not only optimization of drug scheduling (dose-dense approaches) but also use of the drug in combination with other therapies.