Patricia J. Eifel
Radiation therapy plays a major role in the treatment of patients with gynecologic malignancies. For women with cervical cancer, radiation therapy is the primary treatment for patients with advanced disease (1,2), yields cure rates equal to those seen after radical surgery for patients with early tumors (3,4,5), and reduces the risk of local recurrence after surgery for patients with high-risk features (6,7). For women with endometrial cancer, radiation therapy reduces the risk of local recurrence after hysterectomy for patients with high-risk features (8) and is a potentially curative primary treatment for patients with inoperable cancers (9,10). In selected women with ovarian cancer, postoperative, adjuvant, whole-abdominal radiation therapy improves long-term survival (11,12). Radiation therapy is also the primary curative treatment for most patients with invasive vaginal cancer (13,14), and it has an expanding role in the management of carcinomas of the vulva (15,16,17).
Computer technology and information systems have transformed many aspects of radiation-therapy practice in the past two decades, making possible three-dimensional treatment planning based on computed tomography (CT) and magnetic resonance imaging (MRI), optimized inverse planning, computer-controlled treatment delivery, and remote afterloading brachytherapy. These techniques enable radiation oncologists to restrict radiation-dose distributions to specified target volumes, thereby delivering the maximal dose to the tumor, while sparing normal tissues as much as possible.
Radiation biologists and clinicians have also continued to advance our understanding of the molecular mechanisms involved in radiation-induced cell death, the nature of drug-radiation interactions, and the importance of radiation dose, and the time over which it is given, and the dose per fraction. In 1999 and 2000, the results of randomized clinical trials demonstrated a significant improvement in pelvic disease control and survival when concurrent chemotherapy was added to radiation therapy for patients with locally advanced cervical cancer (18,19,20,21). These results led to one of the most significant changes in the standard treatment of gynecologic cancers in decades.
In this chapter, the basic principles of radiation therapy, radiation biology, and radiation physics are reviewed and an overview of the indications for and techniques of radiation therapy in the treatment of gynecologic malignancies is presented.
Radiation Damage and Repair
Cellular Effects of Ionizing Radiation
Cell death can be defined as the loss of clonogenic capacity (i.e., the ability of the cell to reproduce). Most cell death due to ionizing radiation is mitotic cell death. However, ionizing radiation may also cause programmed cell death (apoptosis).
The critical target for most radiation-induced cell death is the DNA within the cell's nucleus. Photons or charged particles interact with intracellular water to produce highly reactive free radicals that in turn interact with DNA to produce strand breaks that interfere with the cell's ability to reproduce. Although this interaction may cause a cell's “reproductive death,” the cell may continue to be metabolically alive for some time. Radiation-induced damage may not be expressed morphologically until days or months later when the cell attempts to divide (mitotic cell death). In some cases, a damaged cell may undergo a limited number of divisions before it dies, having lost the ability to reproduce indefinitely.
Apoptosis (programmed cell death) may also play an important role in radiation-induced cell death (22). In contrast to mitotic cell death apoptosis may occur before cell division or after the cell has completed mitosis. The plasma membrane and nuclear DNA may both be important targets for this type of cell death. Apoptosis appears to be a particularly important mechanism of radiation-induced cell death in certain postmitotic normal tissues, including human salivary glands and lymphocytes. Radiation-induced apoptosis has also been observed in some proliferating normal tissues and tumors. Biologists are actively studying the pathways that regulate the expression of radiation-induced apoptosis in the hope that they can be exploited to improve local tumor control.
The effects of ionizing radiation on the survival of mammalian cell populations in vitro are typically expressed graphically as dose-response or “cell-survival” curves (23). The surviving fraction of cells is plotted (on an exponential scale) against the dose of radiation (on a linear scale). Experimental data using single doses of sparsely ionizing radiation (e.g., x-rays, gamma rays, electrons, or protons) typically produce cell-survival curves with two components (Fig. 4.1): a shoulder region and an exponential region.
Several mathematical models, based on different hypothetical mechanisms of cell killing, have been devised to describe radiation dose-response relationships. These include:
The multitarget model (Fig. 4.1A) is described by the expression logeN = Dq/D0, where N and Dq measure the width of the shoulder and D0 is the slope of the final exponential portion of the survival curve. This model derives from the classic target theory, which holds that each cell contains multiple sensitive targets, all of which must be hit to kill the cell. The presence of a shoulder region is believed to reflect accumulation of sublethal injury in some of the irradiated cells (24). Although the multitarget model accurately describes the exponential portion of the dose-response curve, it is a poor fit to experimental data in the shoulder region. In particular, it fails to predict the approximately linear slope (D1) of the initial portion of the shoulder (Fig. 4.1B).
The linear-quadratic model describes the dose-response relationship according to the equation S = e- α + βD2, where S is the surviving fraction, D is the dose of radiation, and α and β are constants (Fig. 4.1B). This model presupposes two components to cell death: one that is proportional to the dose (αD) and one that is proportional to the square of the dose (βD2). The dose at which the linear and quadratic components are equal is α/β (Fig. 4.1A). This model fits experimental data particularly well for the first few logs of cell death, which are most relevant to fractionated and low-dose-rate (LDR) irradiation, but it is continuously bending on a loglinear plot. This bend is inconsistent with experimental data that demonstrate a straight line on a log-linear plot for the distal portion of the cell-survival curve.
Figure 4.1 Parameters commonly used to characterize the relationship between radiation dose and cell survival in mammalian culture. In the multitarget, or N-D0, model (A), N is the extrapolation number, N and Dq measure the width of the shoulder, and D0 represents the slope of the final exponential portion of the survival curve. The multitarget model provides an accurate description of experimental data in the exponential portion of the survival curve. The linear-quadratic model (B) more accurately describes the shape of the initial shoulder portion of the curve. Because the shoulder has more influence on fractionated radiation therapy, the linear-quadratic model is more often used to predict the results of fractionated clinical radiation therapy. Modified from Hall EJ. Radiobiology for the radiologist, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000, with permission.
Conventional radiation therapy usually is given in a fractionated course with daily doses of 180 to 200 cGy (centiGray) per fraction. Hypothetical cell-survival curves for normal tissue and tumor cells illustrate the advantage of fractionation (Fig. 4.2). When a dose of radiation is divided into multiple smaller doses separated by an interval sufficient to allow maximum repair of sublethal injury, a relatively shallow dose-response curve is achieved, reflecting a repetition of the shoulder of the single-dose cell-survival curve. The slope of the fractionated-dose cellsurvival curve depends on the character of the shoulder (N and Dq). The sparing effect of fractionation is greatest for cells with a response to radiation characterized by a relatively broad shoulder, reflecting the cells' greater ability to accumulate and repair sublethal damage during the interfraction interval. Many normal tissues and some poorly responsive tumors exhibit this type of response to fractionated irradiation in vivo and in vitro. In contrast, most tumors and some acutely responding normal tissues (e.g., bone marrow and intestinal crypt cells) have a dose-response curve with a relatively narrow shoulder, implying relatively little sparing effect of fractionation. The difference between the fractionation sensitivity of tumors and normal tissues is an important determinant of the therapeutic ratio (the difference between tumor control and normal tissue complications) of fractionated irradiation.
So far, this discussion of cell-survival curves and fractionation has referred to radiation given in acute exposures—that is, at a rate of 100 cGy per minute or greater. At these dose rates, the shoulder of the survival curve is pronounced. However, as the dose rate is decreased, cells have a greater opportunity to repair sublethal injury during the exposure. This is called the dose-rate effect. The slope of the survival curve becomes increasingly shallow and the shoulder less apparent (Fig. 4.3) until a dose rate is reached at which all sublethal injury is repaired. In experimental systems, the dose-rate effect appears to be much more pronounced for normal cells than for tumor cells. This differential effect implies a favorable therapeutic ratio that is exploited with LDR intracavitary and interstitial brachytherapy.
The Four Rs
The biological effect of a given dose of radiation is influenced by the dose, fraction size, interfraction interval, and time over which the dose is given. Four factors, classically referred to as “the four Rs of radiobiology,” govern the influence of dose, time, and fractionation on the cellular response to radiation. These are:
As previously discussed, because fractionated irradiation permits greater recovery of sublethal injury during treatment, a higher total dose of radiation is required to achieve a given biological effect when the total dose is divided into smaller fractions. The broader the shoulder of the survival curve, the greater the increase in dose required to achieve the same level of cell death achieved by a single dose. Two-dose experiments with varying interfraction intervals indicate that a space of at least 4 hours, and probably more than 6 hours, is necessary to complete repair of accumulated sublethal injury. Clinical studies tend to confirm these findings; for this reason, altered-fractionation protocols usually require a minimum interval of 4 to 6 hours between treatments.
Repopulation refers to the cell proliferation that occurs during the delivery of radiation. The magnitude of the effect of repopulation on the dose required to produce a given level of cell death depends on the doubling time of the cells involved. For cells with a relatively short doubling time, a significant increase in dose may be required to compensate for a protraction in the delivery time. This phenomenon may be of considerable practical importance. The speed of repopulation of normal tissues that manifest radiation injury soon after exposure (skin, mucosal surfaces, etc.) limits contraction of a course of fractionated irradiation. However, unnecessary protraction probably reduces the effectiveness of a dose of radiation by permitting time for repopulation of malignant clonogens during treatment (25,26). In addition, cytotoxic treatments—including chemotherapy, radiation therapy, and possibly surgical resection—may actually trigger an increase in the proliferation rate of surviving clonogens. This accelerated repopulation may increase the detrimental effect of treatment delays and may influence the effectiveness of sequential multimodality treatments (27,28).
Figure 4.2 Relationship between radiation dose and surviving fraction of cells treated in vitro with radiation delivered in a single dose or in fractions. Top = Most tumors and acutely responding normal tissues. Bottom = Late-responding normal tissues. For most tumors and acutely responding normal tissues, the cellular response to single doses of radiation is described by a curve with a relatively shallow initial shoulder (Top, yellow line). Cellular survival curves for late-responding normal tissues (Bottom, yellow line) have a more pronounced shoulder, suggesting that these cells have a greater capacity to accumulate and repair sublethal radiation injury. When the total dose of radiation is delivered in several smaller fractions (Dose A [dose/fraction] = blue line, or a larger fraction Dose B [dose/fraction] = red line), the response to each fraction is similar and the overall radiation survival curve reflects multiple repetitions of the initial portion of the single-dose survival curve. Note that the total dose required to kill a specific proportion of the cells decreases as the dose per fraction increases (red line). Arrows indicate the differential effects of relatively large versus small fractions of radiation. The greater differential effects of fractionated irradiation on normal tissues (Bottom) than on tumor (Top) reflect the greater capacity of late-responding normal tissues to accumulate and repair sublethal radiation injury. (From Karcher KH, Kogelnik HD, Reinartz G [eds]: Progress in Radio-Oncology II. New York, Raven Press, 1982, pp. 287-296).
Figure 4.3 Response of mouse jejunal crypt cells to different dose rates of α rays. The mice were subjected to total body irradiation, and the proportion of surviving crypt cells was determined by counting regenerating microcolonies in the crypts 3.5 days after irradiation. There was a dramatic difference in cell killing because of repair of sublethal injury at low dose rates. In this system, the lowest dose rate (0.54 cGy per minute) causes little reduction in the number of surviving cells even after high doses because repopulation during the long exposure balances the cell killing from radiation. From Fu KK, Phillips TL, Kane LJ, et al. Tumor and normal tissue response to irradiation in vivo: variation with decreasing dose rates. Radiology 1975;114:709-716, with permission.
Studies of synchronized cell populations have shown significant differences in the radiosensitivity of cells in different phases of the cell cycle (29). Cells are usually most sensitive to radiation in the late G2 phase and during mitosis and are most resistant in the mid- to late S and early G1 phases. When asynchronous dividing cells receive a fractionated dose of radiation, the first fraction tends to synchronize the cells by killing off those in sensitive phases of the cell cycle. Cells remaining in the S phase then begin to progress to a more sensitive phase of the cell cycle during the interval before the next fraction is given. This redistribution of cells to a more sensitive phase of the cell cycle tends to increase the overall cell death achieved from a fractionated dose of ionizing radiation, particularly if the cells have a relatively short cell cycle time.
The sensitivity of fully oxygenated cells to sparsely ionizing radiation is approximately three times that of cells irradiated under anoxic conditions. This makes oxygen the most effective known radiation sensitizer. The molecular interactions responsible for the oxygen effect are not completely understood, but it is believed that oxygen stabilizes the reactive free radicals produced by the ionizing events. The ratio between the dose needed to achieve a given level of cell death under oxygenated versus hypoxic conditions is referred to as the oxygen enhancement ratio (Fig. 4.4).
Most normal tissues are fully oxygenated, but significant hypoxia occurs in at least some solid tumors, rendering the resulting hypoxic cells relatively resistant to the effects of radiation. However, the clinical importance of tumor hypoxia is uncertain because hypoxic cells initially tend to become better oxygenated during a course of fractionated irradiation (30). This phenomenon, called reoxygenation, tends to increase the response of tumors to a dose of fractionated radiation.
Figure 4.4 Survival curves for mammalian cells irradiated under aerated and hypoxic conditions. The dose required to produce a given level of damage is approximately three times greater under hypoxic or anoxic conditions than under fully oxygenated conditions. The ratio of doses is the oxygen enhancement ratio (OER). Sometimes the shoulder also is reduced under hypoxic conditions. Modified from Hall EJ. Radiobiology for the radiologist, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000, with permission.
Treatment Strategies for Overcoming Radioresistance of Hypoxic Cells
Many treatment strategies have been explored to overcome the relative radioresistance of hypoxic cells in human solid tumors (31,32,33,34,35). These include:
None of these approaches has clearly demonstrated an improvement in outcome; however, most of the studies have been severely compromised by technical or logistical problems.
Numerous retrospective studies have found correlations between the minimum hemoglobin level during treatment and outcome, but all of them have been compromised by possible confounding risk factors (36,37,38). Even with multivariate analysis, investigators cannot rule out the possibility that patients whose hemoglobin levels fell despite transfusion also had tumors that were more aggressive or less responsive to treatment. Studies of intratumoral oxygen tension also suggest that hypoxic tumors tend to have a poor prognosis; however, this correlation appears to be present even in surgically treated patients and may in part reflect a tendency for biologically aggressive tumors to be hypoxic (39).
An early randomized study of transfusion in anemic patients with locally advanced cervical cancer (40) hinted at improved local control when oxygen-carrying capacity was increased. However, the findings of this small study have not yet been confirmed in a larger prospective trial, and the results remain inconclusive. One group of investigators (41) has even suggested that allogeneic transfusion may be harmful, although their results conflict with those of most other studies. Nevertheless, tumor hypoxia continues to be one probable cause of the failure of irradiation to control some tumors (e.g., advanced cervical cancers with a significant population of hypoxic tumor cells), and most clinicians recommend that the hemoglobin level be maintained above 10 g/dl during radiation therapy (42).
In the late 1990s, the commercial availability of recombinant erythropoietin led investigators to explore the impact of growth-factor-induced increases in hemoglobin level on the outcome of patients treated with radiation therapy. Initial enthusiasm was tempered by negative results of a randomized trial in patients with head and neck cancer and by reports of increased thromboembolic events in patients receiving erythropoietin (43). The Gynecologic Oncology Group prematurely closed a randomized trial of chemoradiation with or without erythropoietin in patients with locally advanced cervical cancer because of concerns about the risk of thromboembolism (44). In that study, thrombotic events occurred in 11 of 57 (19.3%) who received erythropoietin versus four of 52 (7.7%) treated with chemoradiation alone (p = NS); the impact of erythropoietin on outcome was inconclusive because of the small number of patients in the study.
Linear-Energy Transfer and Relative Biological Effectiveness
The rate of deposition of energy along the path of the radiation beam is referred to as its linear-energy transfer (45). Photons, high-energy electrons, and protons produce sparsely ionizing radiation beams (low linear-energy-transfer), whereas larger atomic particles (e.g., neutrons and alpha particles) produce much more densely ionizing radiation beams (high linear-energy transfer). The biological effects of densely ionizing radiation beams differ in several important ways from those of more sparsely ionizing radiation beams. With highlinear-energy-transfer radiation beams:
The unit of relative biological effectiveness is used to compare the effects of different radiation beams. Relative biological effectiveness is defined as the ratio between a test radiation dose and that of 250-kV x-rays needed to produce a specific biological effect. The relative biological effectiveness may differ somewhat according to the tissue and biological end point being studied.
In practice, few facilities exist for the production of high-linear-energy-transfer beams, and their use has had no major impact on the results of treatment for gynecologic malignancies.
Temperature is another factor that can modify the effect of ionizing radiation (46). Supraphysiological temperatures alone can be toxic to cells because heat is preferentially toxic to cells in a low-pH environment (frequent in areas of hypoxia) and to cells in the relatively radioresistant S phase of the cell cycle. Temperatures in the range of 42° to 43°C sensitize cells to radiation both by reducing the shoulder and by increasing the slope of the cellsurvival curve. Because of the different vascular supplies of tumors and normal tissues, hyperthermia may produce greater temperature elevations in tumors, increasing the possible therapeutic advantage when heat is combined with irradiation. Biologists and clinicians have been trying to find ways to exploit this effect for many years but have been hampered by technological limitations on the ability to selectively heat deep-seated tumors (47). A trial from Amsterdam (48) reported that survival was improved when hyperthermia was used with irradiation in patients with locally advanced cervical cancer. The patients in this study received relatively low doses of radiation, did not receive concurrent chemotherapy, and had poorer than expected pelvic disease control in the control arm, but the findings suggest that the approach may still deserve further study.
Interactions between Radiation and Drugs
Drugs and radiation interact in a number of ways to modify cellular responses. Steel and Peckham (49) categorized these interactions into four groups: spatial cooperation (independent action), additivity, supraadditivity, and subadditivity.
Spatial Cooperation (Independent Action)
Spatial cooperation is the situation in which drugs and radiation act independently with different targets and mechanisms of action so that the total effect of the combination is equal to that of each agent separately. For example, a site that is protected from chemotherapy (e.g., the brain) may be treated with radiation to prevent recurrence. Alternatively, a drug may be used to destroy microscopic distant disease while radiation is used to sterilize local tumor.
Additivity is the situation in which two agents act on the same target to cause damage that is equal to the sum of their individual toxic effects.
When there is supraadditivity, a drug potentiates the effect of radiation, causing a greater response than would be expected from simple additivity.
With subadditivity, the amount of cell death that results from the use of two agents is less than that expected from simple additivity (the amount may still be greater than expected from either treatment alone).
Clinically, it is difficult to determine which mode of interaction occurs when two agents are used concurrently. When a greater response is observed than would be expected from radiation alone, the interaction is often described as synergistic but may be only additive or even subadditive.
Ionizing radiation interacts with all the tissues in its path, not just tumor tissue. Radiation can be considered an effective cancer treatment only if there is a differential biological effect on tumor and normal tissues. The difference between tumor control and normal tissue complications is referred to as the therapeutic gain or therapeutic ratio.
In general, the relationship between the probability of tumor cure or the probability of normal tissue injury and the dose of radiation can be described by a sigmoid curve (Fig. 4.5). At relatively low radiation doses, there is an insufficient amount of cell death to produce any likelihood of tumor cure. As the dose is increased, a threshold is reached at which some cures begin to be observed. For most tumor systems, the likelihood of cure rises rapidly as the radiation dose is increased beyond this threshold and then reaches a plateau. The shape and slope of the dose-response curve vary according to the tumor type and size (50, 51). A similar sigmoid relationship is seen when the likelihood of complications is plotted against the radiation dose. If the sigmoid curve for normal tissue complications is to the right of the sigmoid curve for tumor control, then treatment with doses that fall between the two curves may achieve tumor control without causing complications. The difference between these curves represents the therapeutic ratio. The primary goal of radiation research efforts is to improve the therapeutic ratio by increasing the separation between these dose-response curves, maximizing the probability of complication-free tumor control.
Figure 4.5 Theoretical sigmoid dose-response curves for tumor control and severe complications. The therapeutic ratio is related to the distance between the two curves. Dose A controls tumor in 80% of cases with a 5% incidence of complications. Dose B yields a 10% to 15% increase in the tumor control probability but a much greater risk of complications, narrowing the therapeutic ratio. A leftward shift of the tumor control probability curve (e.g., by the addition of sensitizing drugs) broadens the window for complication-free cure.
Effects of Radiation on Normal Tissues
The extent of radiation damage to normal tissues depends on a number of factors, including the radiation dose, the organ, the volume of tissue irradiated, and the division rate of the irradiated cells. Tissues that have rapid cell turnover (i.e., tissues whose functional activity requires constant cell renewal) tend to manifest radiation injury soon after exposure, often during a fractionated course of radiation therapy. Examples of acutely reacting tissues include most epithelia (e.g., skin, hair, gastrointestinal mucosa, bone marrow, and reproductive tissues). In contrast, tissues that have slower cell turnover (i.e., tissues whose functional activity does not require constant cell renewal) tend to manifest radiation injury months or years after exposure to radiation. Examples of late-reacting tissues are the connective tissues, muscle, and neural tissues. In some normal tissues, cell death may occur through the mechanism of apoptosis. Although apoptosis is not the primary mechanism of damage in most normal tissue injury, it is important in the response of lymphocytes, salivary gland cells, and a small proportion of intestinal crypt cells (22).
Acute reactions to pelvic irradiation, such as diarrhea, are usually associated with mucosal denudation, which in turn stimulates an increase in cell proliferation (52). This regenerative response is usually sufficient to prevent serious side effects with weekly doses of 900 to 1,000 cGy given in five fractions. This empirically derived schedule is the most commonly used for clinical radiation therapy. If treatment is accelerated to deliver the dose over much shorter periods, then the regenerative capacity of the epithelium may be overwhelmed and the acute reaction so severe that a break in treatment is needed to allow for epithelial regeneration. The severity of acute reactions also depends on the volume of the normal tissues irradiated and the specific nature of the tissues.
The pathogenesis of late radiation complications (i.e., those that occur months to years after radiation therapy) differs from that of acute reactions and is still incompletely understood. It has been hypothesized that late effects of radiation result from:
Because late-reacting tissues are not proliferating rapidly, the duration of a course of radiation treatment does not alter their tolerance. However, late-responding normal tissues tend to be quite sensitive to changes in the dose per fraction so that a strong correlation is seen between the radiation fraction size and the risk of late complications. Thus, for a given dose of radiation administered over a given period, the risk of late effects will be greater with larger fractions. This fractionation effect is responsible for the advantage of altered fractionation schedules in clinical settings in which late normal tissue reactions are severely dose limiting (53,54,55) (Fig. 4.2). This fractionation effect also has important implications for treatments such as high-dose-rate (HDR) brachytherapy and intensity-modulated radiation therapy (IMRT), where a portion of the target (and possibly adjacent normal tissues) frequently receives doses of more than 2 Gy per fraction.
The likelihood of developing serious late effects from radiation depends on many factors, including but not limited to the dose of radiation, the radiation dose per fraction, the volume of tissue irradiated, the radiation-dose rate, patient characteristics, other treatments (such as surgery or chemotherapy), and the end point being measured. Some tissues—such as the liver, kidney, and lung—consist of functional subunits that are arranged more or less in parallel; these tissues can tolerate a high dose of radiation given to a small portion of the organ without serious late effects but tend to be relatively sensitive to moderate whole-organ doses. Other organs, such as bowel or ureter, are organized in a serial fashion—delivery of a damaging dose to even a small portion of the organ can cause total organ failure. For all of the reasons discussed previously, normal tissue tolerances cannot be described in terms of simple dose limits. However, some generalizations can be made about the tolerance of individual tissues (doses refer to external radiation given in daily fractions of 1.8 to 2 Gy or with LDR brachytherapy).
Uterus The uterus and cervix are typically described as resistant to radiation; however, what is really meant by this is that the uterus can be treated to very high doses (more than 100 Gy in some cases) without the patient developing serious complications in adjacent critical structures (e.g., bowel and bladder). The uterus probably cannot sustain pregnancy after such doses. Even moderate doses of 40 to 50 Gy probably cause enough smooth muscle atrophy to prohibit successful term pregnancy, but this has rarely been tested. Women who received 20 to 30 Gy or more to the uterus during the perimenarchal period have become pregnant but tend to have spontaneous second-trimester abortions, probably because of underdevelopment of the uterus. Patches of endometrium frequently continue to function after doses of 50 Gy or more.
Ovary The radiation dose required to cause ovarian failure is highly dependent on the patient's age. Perimenarchal girls may continue to menstruate and can even become pregnant after receiving as much as 30 Gy to the ovaries; however, they usually experience premature menopause 10 to 20 years later. Most adult women have ovarian failure after 20 Gy; as little as 5 to 10 Gy can induce menopause in older premenopausal women.
Vagina The radiation tolerance of the vagina depends on the region (upper, mid-, lower, anterior, posterior, or lateral) and length of vagina treated as well as radiation dose, fraction size, dose rate, hormonal support, and other factors. Small portions of the surface of the lateral apical vagina can be treated to a very high dose (≥140 Gy) without causing major complications in adjacent structures. However, these high doses do cause atrophy and shortening of the apical vagina. The vaginal tolerance dose is less if treatment includes more than the apical vagina or if the dose includes the posterior, or distal vagina. Even moderate doses (40 to 50 Gy) may decrease the elasticity of the vagina, although it is sometimes difficult to distinguish the direct effects of radiation from those of tumor, altered hormonal environment, aging, and other factors.
Small Intestine The risk of small intestinal side effects is highly dependent on the radiation dose and volume irradiated and on the patient's history. In the absence of complicating factors, the entire small intestine can tolerate doses of as much as 30 Gy without major late effects. Smaller volumes can tolerate 45 to 50 Gy with a low risk of complications; the risk of chronic diarrhea and bowel obstruction increases rapidly with doses greater than 60 Gy and approaches 100% if a significant volume of small bowel receives 70 Gy or more. The risk of bowel obstruction is significantly increased in patients who have a history of major transperitoneal surgery, pelvic infection, or heavy smoking (56).
Rectum In most cases, the entire rectum can tolerate 45 to 50 Gy with a low risk of major sequelae. Small portions of the anterior rectal wall can tolerate doses of at least 70 to 75 Gy. However, the risk of serious late effects (severe bleeding, obstruction, fistula) increases steeply as the volume of rectum treated to high dose is increased.
Bladder The entire bladder can be treated to 45 to 50 Gy with a very low rate of serious morbidity. This dose may have subtle effects on bladder contractility, particularly in patients who have also undergone radical hysterectomy. Small portions of the bladder tolerate doses of 80 Gy or more with a low risk of major morbidity (severe bleeding, contracture, fistula). However, the dose-response relationship is poorly defined in this range because it is difficult to accurately determine the maximum dose given to the bladder during intracavitary treatments.
Ureter Surgically undisturbed ureters appear to tolerate 85 to 90 Gy of combined external-beam and LDR intracavitary treatment with a low risk of stricture.
Kidneys Most patients can tolerate as much as 18 to 22 Gy to both kidneys with very little risk of long-term damage. Higher doses cause permanent damage to renal parenchyma. If the patient has normal renal function, then 50% or more of the renal parenchyma can be treated to a high dose without causing renal failure; however, renal hypertension may occur if an entire kidney is obliterated with radiation. Underlying renal disease or concurrent use of chemotherapy can decrease renal tolerance.
Liver In most cases, the liver can tolerate as much as 30 Gy (at 1.5 Gy per fraction) to the entire organ, although this dose will cause transient elevation of alkaline phosphatase levels and can cause dysfunction in a small proportion of patients. Higher doses cause serious damage to liver parenchyma but can be tolerated if delivered to a portion of the liver only. Tolerance is highly dependent on underlying hepatic function and can be markedly decreased with concurrent delivery of some chemotherapeutic agents and during periods of hepatocyte regeneration (for example, after partial hepatectomy).
Spinal Cord and Nerves Transverse myelitis and paralysis can occur in a small proportion of patients who receive doses as low as 50 Gy to the spinal cord, and the risk increases rapidly as the dose approaches 60 Gy at 2 Gy per fraction. However, peripheral nerves, including the cauda equina, are rarely affected after 50 Gy and usually tolerate doses as high as 60 Gy without serious sequelae.
Bone As little as 10 to 15 Gy of radiation causes transient depletion of bone marrow elements. With doses of more than 30 to 40 Gy, permanent damage is done to supporting elements, and bone marrow within the irradiated area will not repopulate normally. This damage can be seen as fatty replacement of the marrow cavity on MRI. The risk of fracture after radiation therapy depends on the bone irradiated, the volume of bone in the high-dose region, bone density, coexistent steroid use, and other factors. Symptomatic fracture is rare after treatment with 40 to 45 Gy of pelvic radiation. However, routine MRI sometimes detects small, usually asymptomatic insufficiency fractures of the pelvis after this dose (57). Hip fracture may be seen after doses as low as 40 Gy to the entire femoral head and neck, and the risk probably increases rapidly as the dose approaches 60 Gy.
Treatment Strategies to Exploit Differences between Tumor and Normal Tissue in the Response to Fractionated Radiation Therapy
A variety of altered fractionation schemes have been devised to exploit the different sensitivities of tumor and normal tissues to fractionation and the possible effects of tumor cell repopulation. These include hyperfractionation, in which the dose per fraction is reduced, the number of fractions and total dose are increased, and the overall treatment time is relatively unchanged; accelerated fractionation, in which the dose per fraction is unchanged, the overall treatment duration is reduced, and the total dose is unchanged or decreased; and hypofractionation, in which the dose per fraction is increased, the number of fractions and total dose are reduced, and the overall treatment time is decreased.
With hyperfractionation, treatment is usually given two or more times daily with at least 4 to 6 hours between fractions to allow repair of sublethal injury. This scheme should permit delivery of a higher dose of radiation without increasing the risk of late complications or the overall duration of treatment. Hyperfractionation schemes may have an advantage if the increased dose delivered per day does not cause unacceptable acute effects and if patients are willing to accept the added inconvenience of two or three treatments daily.
Accelerated fractionation schemes do not reduce the risk of late effects and tend to increase the acute effects of treatment but may be advantageous because treatment is completed over a shorter time, reducing tumor cell repopulation during treatment (55). However, such schemes are likely to be of limited value in the management of gynecologic malignancies because acute side effects tend to limit the rate of treatment delivery.
Hypofractionation schedules are usually avoided when treatment is likely to cure the patient because the α/β of late-responding normal tissues is less than the α/β of most tumors, meaning that large fractions have a therapeutic disadvantage. Malignant melanoma, which appears to have a relatively low α/β, may be a rare exception to this pattern. Hypofractionated schedules are frequently used for palliative treatment because they are convenient and produce rapid symptom relief. However, the necessary reduction in dose reduces the likelihood of complete eradication of tumor within the treatment field. Hypofractionation may be particularly beneficial if the tumor target is some distance from critical structures and if the radiation treatment plan is characterized by a rapid dose gradient such that the target receives a relatively high dose per fraction while normal tissue structures receive no more than approximately 2 Gy per fraction. Under ideal circumstances, HDR brachytherapy plans and some highly conformal external-beam plans achieve this favorable geometry.
Combinations of Surgery and Radiation Therapy
Because surgery and radiation therapy are both effective treatments, clinicians have tried to improve locoregional control or reduce treatment morbidity by combining the two modalities. Theoretically, surgery may remove bulky tumor that may be difficult to control with tolerable doses of radiation, and radiation may sterilize microscopic disease at the periphery of the surgical bed. The two modalities have been combined in a number of ways:
Preoperative irradiation is sometimes used to sterilize possible microscopic disease at the margins of a planned operative site. This is potentially most useful when the surgeon anticipates close margins adjacent to a critical structure—for example, the urethra or anus in a patient with locally advanced vulvar cancer.
Preoperative irradiation has largely been abandoned in favor of postoperative irradiation, which can be planned when information from the surgical specimen is available and which avoids unnecessarily treating patients with very-early-stage disease. Preoperative irradiation is still sometimes used to treat patients with stage II endometrial cancer that grossly involves the cervix and is also used in some patients with bulky cervical cancers. This is because the dose deliverable to paravaginal tissues is much greater when the uterus is still in place to hold an intrauterine applicator than after surgery, when only an intravaginal applicator can be used.
Some studies have suggested that lower doses of radiation may be required to sterilize microscopic disease in a tumor bed undisturbed by surgery because an intact vascular supply is better able to deliver oxygen. Because the risk of operative complications is increased after high-dose radiation, doses given when surgical resection is anticipated are usually lower than doses given when a tumor is irradiated definitively. The greatest risk of preoperative radiation therapy is that if the tumor remains unresectable, the effectiveness of additional irradiation will be markedly decreased by the long interval between treatments.
In some cases, intraoperative irradiation can be delivered with a permanent implant (using 125I or 198Au), with afterloading catheters in the operative bed (using 192Ir), or with a special electron beam or orthovoltage unit in the operating room. These approaches deliver radiation directly to the site of maximum risk when the target can be visualized directly and normal tissues nearest the treatment area can be removed from the radiation field. Removal of normal tissues from the treatment field is an important physical advantage of intraoperative external-beam techniques that must counterbalance the biological disadvantage to any normal tissues remaining in the field when an entire dose is delivered in a single large fraction.
Postoperative irradiation has been demonstrated to improve locoregional control and even survival in several settings important to gynecologic oncologists. In vulvar cancer,postoperative pelvic and groin irradiation reduces the risk of groin recurrence and improves the survival rate of patients with multiple positive inguinal nodes (16). In endometrial cancer, postoperative pelvic irradiation reduces the incidence of pelvic recurrences in patients with high-risk disease (8,58,59). In cervical cancer, postoperative pelvic irradiation reduces pelvic recurrence in patients with lymph node involvement and in those with high-risk features in the primary tumor (6,7).
Combined therapy is optimized when the treatment plan exploits the complementary advantages of the two treatments. This requires close cooperation between specialists at the time of the patient's initial evaluation. Because the morbidity of combined therapy is often greater than that of single-modality therapy, combined treatment should usually be limited to situations in which a combined approach is likely to improve survival, permit organ preservation, or significantly reduce the risk of local recurrence compared with the expected results from treatment with either modality alone (60).
Ionizing Radiations Used in Therapy
Ionizing radiations lie on the high-energy portion of the electromagnetic spectrum and are characterized by their ability to excite, or ionize, atoms in an absorbing material. The nuclear decay of radioactive nuclei can produce several types of radiation, including uncharged gamma (γ) rays, negatively charged beta (β) rays (electrons), positively charged alpha (α) particles (helium ions), and neutrons. The resulting ionizing radiations are exploited therapeutically in brachytherapy treatments (using 226Ra, 137Cs, 186Ir, and other isotopes) or to produce teletherapy beams (e.g., 60Co). The average energy of the photons produced by the decay of radioactive cobalt is 1.2 million eV (MeV).
Most external-beam therapy is delivered via linear accelerators that produce photon beams (x-rays) by bombarding a target such as tungsten with accelerated electrons. Varying the energy of the accelerated electrons produces therapeutic x-rays of different energies. X-rays and γ-rays are both composed of photons and differ only in that x-rays are produced by extranuclear forces and γ-rays are produced by intranuclear forces.
Interactions of Radiation with Matter
X-Rays and γ-Rays
Photons interact with matter by means of three distinct mechanisms: the photoelectric effect, Compton scatter, and pair production.
The photoelectric effect is most important at energies used for diagnostic purposes. Absorption by the photoelectric effect is proportional to Z3, where Z is the atomic number of the absorbing material. This effect is responsible for the increased absorption of bone that provides contrast between bone and soft tissue with diagnostic x-ray beams of 250 kV or less. However, the increased bone absorption, high skin dose, and poor penetration with such beams make these beams unsuitable for most modern therapeutic applications. Superficial kilovoltage radiation beams, delivered using a transvaginal cone, are occasionally used for patients with large bleeding exophytic tumors to achieve hemostasis before definitive treatment (61).
Modern therapeutic beams of 1 to 20 megavolts (MV) produce photons that interact with tissues primarily by Compton scatter. In this process, incident photons interact with loosely bound outer-shell electrons, ejecting them from the atom. Both the photon and the electron go on to interact with other atoms, causing additional ionizations. Compton-scatter absorption is independent of Z but varies according to the density of the absorbing material. This accounts for the poor contrast of radiation portal verification films.
Photons that are absorbed by Compton scatter produce an increasing number of scattered electrons and ionizations as they penetrate beneath the surface of an absorbing material. This creates a buildup region just below the surface that is responsible for the skin-sparing characteristic of modern high-energy therapy beams (Fig. 4.6). The maximum dose from a megavoltage beam is reached at 0.5 to 3.0 cm below the skin surface, depending on the photon energy. At greater depths, the dose decreases at a fairly constant rate that is related to the beam energy. The greater skin-sparing effects and penetration of energy beams of 15 MV or greater make them particularly useful for pelvic treatment.
Pair production absorption is related to Z2. In soft tissue, this type of absorption begins to dominate only at photon energies of more than approximately 30 MeV, so pair production is of limited importance in current radiation therapy planning.
Electrons and Other Particles
Several types of particle beams are used in radiation therapy: electron beams, proton beams, and neutron beams.
Electrons are very light particles. When they interact with matter, they tend to lose most of their energy in a single interaction. The dose from an electron beam is relatively homogenous up to a depth that is related to the beam's energy (Fig. 4.6). Beyond this depth, the dose decreases very rapidly to nearly zero. Electrons are used to treat relatively superficial targets without delivering a significant dose to underlying tissues. The approximate depth (in centimeters) at which the rapid falloff in dose occurs can be estimated by dividing the electron energy by 3.
Protons are positively charged particles that are much heavier than electrons. Protons scatter minimally as they interact with matter, deposit increasing amounts of energy as they slow down, and then stop at a depth related to their initial energy. This results in rapid deposition of most of their energy at depth (called the Bragg peak), with a steep falloff in dose to near zero shortly after the peak. Modulating the energy can spread this peak out. The absence of an exit dose makes proton beams ideal for conformal therapy, and interest in their use has increased as the cost of producing proton generators has become somewhat more reasonable. The physics support, quality assurance, and clinical requirements needed to safely treat patients with protons are complex, highly specialized, and time consuming. Although in some difficult clinical situations, protons clearly provide at least a theoretical dosimetric advantage over photons, there are as yet no randomized comparisons. Also, because the depth of penetration of protons is highly dependent on the density of intervening tissue, the presence of variable gas-filled structures (e.g., bowel) in the midpelvis may limit applications in gynecologic oncology.
Figure 4.6 Depth dose curves for selected x-ray and γ-ray beams (top). As the energy increases, the depth of maximum dose (Dmax or D100) increases. For kilovoltage beams, the dose is maximum at the skin surface. With appositionally directed megavoltage beams (e.g., 60Co or 25-MeV photon beams), the maximum dose is reached at a depth beyond the skin surface, producing skin sparing. High-energy beams also penetrate more deeply, making them more useful for treatment of deep-seated pelvic tumors.
Depth dose curves for electron beam fields of selected energies (bottom). The depth of maximum dose increases with increasing energy. At depths just below the maximum, the dose falls off rapidly, sparing deeper tissues.
Neutrons are neutral particles that tend to deposit most of their energy in a single intranuclear event. For this reason, there is little or no repairable injury and therefore no shoulder on the tumor cell-survival curve. The falloff of a neutron dose is similar to that of a photon beam of 4 to 6 MV, but the high relative biological effectiveness of densely ionizing neutron beams has been of interest to clinical investigators. However, clinical studies of neutron treatments in cervical cancer patients were plagued by high complication rates (62), and neutrons are rarely if ever used to treat gynecologic tumors today.
Measurement of Absorbed Dose
Absorbed dose is a measure of the energy deposited by the radiation source in the target material. The unit currently used to measure radiation dose is the Gray (Gy), equal to 1 joule per kilogram of absorbing material. Before the early 1980s, absorbed doses of radiation were measured in radians (rads), where 1 rad = 1 cGy and 1 Gy = 100 rad.
The rate of decay of a sample of radioactive material (such as radium or cesium) is referred to as the activity of the sample and is measured in curies (Ci), where 1 Ci = 3.7 × 1010disintegrations per second, and 1 mCi = 10-3 Ci.
Safe delivery of radiation depends on precise calibration of radiation source activities and machine output. These are measured using sensitive ionization chambers in phantoms that simulate tissue density. Periodic calibrations of equipment and sources are a vital part of quality assurance in any radiation oncology department.
Inverse Square Law
The dose of radiation from a source to any point in space varies according to the inverse of the square of the distance from the source to the point (63). This relationship is particularly important for brachytherapy applications because it results in a rapid falloff of dose as distance from an intracavitary or interstitial source is increased.
Radiation therapy is delivered in three ways.
Several terms are commonly used to describe the dose distributions produced by external-beam irradiation of tissues.
Percentage depth dose is the change in dose with depth along the central axis of a radiation beam (Fig. 4.6).
Dmax is the maximum dose delivered to the treated tissue. With a single appositional photon beam, the Dmax is located at a distance below the tissue surface that increases with the energy of the photon beam (Fig. 4.6).
Source to skin distance is the distance between the source of x-rays (e.g., a cobalt source or the target in a linear accelerator) and the skin surface.
Isocenter is a point within the patient that remains a fixed distance from the radiation source as the treatment source (gantry) is rotated around the patient (Fig. 4.7).
Source to axis distance is the distance from the source of x-rays to the isocenter.
Isodose curve is a line or surface that connects points of equal radiation dose (Fig. 4.8).
Figure 4.7 Diagram of a therapeutic linear accelerator. Patients are positioned on the treatment couch with a system of lasers that are aligned precisely with the center of the radiation beam. Collimators in the treatment head, located on a rotating gantry, define the size and rotation of the radiation field. The treatment couch can also be rotated around the central axis of the radiation beam. Beam-modifying devices such as shielding blocks and wedges can be attached to a tray beneath the collimator (not shown).
Figure 4.8 Isodose distribution for external-beam irradiation of the pelvis using an 18-MV beam. (A) A pair of parallel opposed anterior and posterior fields. (B) Anterior, posterior, and two lateral fields (four-field box technique). The heavy red isodose line represents the region of tissue treated to ≥ 45 Gy.
Many factors influence the dose distribution in tissue from a single external beam of photons. These include:
Modern linear accelerators permit many variations in these factors (Fig. 4.7). A rotational gantry permits isocentric beam arrangements that maintain a fixed distance between the beam's source and a point within the patient. This facilitates accurate patient setup and treatment planning.
Most radiation therapy treatment plans combine two or more beams to create a dose distribution designed to accomplish three aims: (i) to maximize the dose of radiation delivered to the target; (ii) to produce a relatively homogeneous dose within the volume of interest to minimize hot or cold spots that would increase the risks of complications or recurrence, respectively; and (iii) to minimize the dose delivered to uninvolved tissues, taking into account the different tolerances of various normal tissues.
The treatment plan must include the primary target volume (gross tumor or tumor bed), any areas at risk for microscopic spread of disease, and a margin of tissue to account for uncertainties in the location of the target, reproducibility of the setup, and organ motion. The overall plan is often designed to deliver different doses to areas of greater or lesser risk (e.g., gross versus microscopic residual disease) by boosting areas at greater risk with smaller treatment fields after initial treatment to a relatively large volume. Two opposing beams (e.g., anterior-posterior and posterior-anterior) usually produce a relatively homogeneous distribution of dose within the intervening tissue with some sparing of the skin surface. In many cases, multiple fields may be used to “focus” the high-dose region to conform more closely to a deep target volume (Fig. 4.8).
Modern technology has made it possible to use computers to optimize the beam arrangements required in treatment plans that incorporate many fields and beam-shaping devices. These conformal treatment plans may provide a very tight distribution of dose around the target volume. The simplest form of conformal therapy uses fairly conventional beam arrangements but exploits modern CT-based treatment-planning techniques to more accurately define the target volume and to design blocks that conform closely to that volume. CT reconstructions permit more accurate shaping of fields that enter the patient from oblique angles. Multileaf collimators have computer-controlled leaves that can form irregularly shaped fields, replacing hand-loaded beam-shaping devices. Because the therapist no longer needs to enter the room to replace blocks on each field, it is possible to treat patients with more fields and more complex beam arrangements in a treatment visit of acceptable duration.
More recently, attention has been focused on IMRT (Fig. 4.9). This form of highly conformal radiation therapy uses complex computer algorithms to optimize delivery of radiation from multiple beam angles. The physician must carefully contour target volumes and all critical normal tissue structures on each slice of a CT scan that has been obtained in the treatment position. The minimum and maximum acceptable doses of radiation to be delivered to each area are specified. Inverse planning techniques (based on the physician's designation of targets and avoidance structures rather than specific radiation fields) are used to design an optimized plan, which usually includes multiple irregularly shaped fields from each of several (usually six to nine) beam angles.
Figure 4.9 Dose distribution obtained using intensity-modulated radiation therapy (IMRT) to treat the pelvic lymph nodes after hysterectomy. In this case, each of seven fields was modulated to obtain a distribution that covered the iliac and presacral lymph nodes while sparing bowel in the central pelvis from high dose. A somewhat larger volume receives low-dose radiation than with standard techniques, and the very tight dose distribution requires an accurate understanding of anatomy, tissues at risk, and internal organ motion.
The leaves of multileaf collimators enter the field or retract dynamically to deliver the desired amount of radiation to tissues within the target. Very tightly conforming radiation distributions can be obtained with this approach. However, the time required to plan treatments is lengthened, as is the duration of daily treatments. Quality assurance is also very demanding for this type of treatment because the fields are less readily visualized than static radiation fields. In the past 5 years, the use of IMRT and other highly conformal radiation techniques has increased exponentially. In many cases, these techniques clearly can be used to reduce the dose delivered to normal tissues during a course of radiation therapy. However, the opportunities for error also are increased; unlike traditional treatments that were based on relatively simple, empirically tested field shapes and distributions, IMRT plans are entirely dependent on the clinician's understanding of the target volume and tissues at risk. If the clinician misses or fails to correctly designate tissues at risk for disease, the computerized inverse planning process will tend to result in exclusion of areas of possible tumor involvement or overtreatment of critical structures. Because the dose of radiation falls off rapidly outside the designated target volume, IMRT plans require a high degree of confidence in the distribution of disease, a clear understanding of internal organ motion, and meticulous patient immobilization. Because there are as yet no level-1 data confirming the benefit of IMRT in treatment of gynecologic neoplasms,payers may consider the treatment experimental and decline payment.
Any treatment that involves placement of radioactive sources within an existing body cavity is termed intracavitary treatment. The most common gynecologic applications of intracavitary therapy involve placement of intrauterine or intravaginal applicators that are subsequently loaded with encapsulated radioactive sources (e.g., 137Cs or 192Ir) (Table 4.1). Applicator systems vary in their appearance and configuration, but those used for radical treatment of cervical or uterine cancer tend to have several features in common. These applicators usually consist of a hollow tube, or tandem, and some form of intravaginal receptacle for additional sources. The greatest variation between systems is in the vaginal applicators, which differ in their shape, the orientation of sources, and the presence or absence of shielding (64,65). One applicator that is commonly used to treat intact carcinomas of the cervix is the Fletcher-Suit-Delclos system. Important characteristics of this system are the arrangement of vaginal sources perpendicular to the tandem and the presence of internal shielding that reduces the dose to the rectum from the vaginal sources by as much as 25%. The Fletcher-Williamson applicator (Fig. 4.10) is similar to the Fletcher-Suit-Delclos applicator but is adapted for use with an iridium stepping source (66).
Table 4.1 Isotopes Used in Gynecologic Oncology
The vaginal ring applicator is commonly used with high dose rate HDR systems and has geometry similar to that of the unshielded Delclos miniovoids used with Fletcher-type applicator systems. Other applicator systems, such as the Delclos dome cylinder, have been designed specifically for treatment of the vaginal apex after hysterectomy (67).
Figure 4.11 illustrates a typical pear-shaped isodose distribution produced by a line of intrauterine sources and Fletcher-Suit-Delclos vaginal colpostats loaded with 137Cs.Intracavitary brachytherapy has proven very useful in the treatment of cervical cancer because it allows a very high dose of radiation to be delivered to a small volume surrounding the applicator (i.e., the cervix and paracervical tissues) without excessive treatment of normal tissues that are more distant from the sources. Because of the rapid change in dose over short distances, accurate positioning of the intracavitary applicator and sources is very important. Packing or retraction of the bladder and rectum can significantly reduce the dose to portions of these organs by distancing them from the vaginal sources.
Figure 4.10 Fletcher-Williamson type applicators used for high-dose-rate and pulsed doserate intracavitary applications. Note the tungsten shields located in the inferior-medial position anteriorly and posteriorly in the small (2 cm) ovoid inserts.
Figure 4.11 Posterior-anterior and lateral views of a Fletcher-Suit-Delclos applicator system loaded with 137Cs sources for treatment of invasive cervical cancer. Units on the isodose contours are cGy per hour. Point A (A), bladder (B), and rectal (R) reference points are indicated on the figure. From Eifel PJ, Berek JS, Thigpen JT.Cancer of the cervix, vagina, and vulva. In DeVita V, Hellman S, Rosenberg S, eds. Cancer: principles and practice of oncology. Philadelphia: JB Lippincott Co., 2001;1526-1556, with permission.
To minimize the exposure of medical personnel to radiation, most modern applicator systems are loaded with radioactive sources after adequate positioning has been confirmed with anterior-posterior and lateral x-rays of the pelvis. In many cases, remote afterloading devices are used to automatically retract sources from the applicator to a lead-lined safe when someone enters the patient's room, further reducing the radiation exposure to visitors and medical personnel.
Historically, most brachytherapy was delivered at a low dose rate, most commonly 40 to 60 cGy per hour. These dose rates take maximum advantage of the dose-rate effect described above, differentially sparing late-responding normal tissues as compared with acutely responding tissues and tumor cells. The dose of LDR intracavitary therapy needed to radically treat cervical cancer is usually delivered in 72 to 96 hours during one or two hospital admissions. Although some investigators have tried to reduce the duration of these treatments by doubling the dose rate (from 40 cGy per hour to 80 cGy per hour), the limited clinical data on this approach suggest that doubling the dose rate results in a less favorable therapeutic ratio (68).
In the past two decades, the advent of computer-controlled remote afterloading has made it possible to deliver brachytherapy treatments at high dose rates (in minutes rather than hours). HDR treatment may offer practical advantages for the patient because it is typically performed on an outpatient basis, although more applications are usually required. With this technique, a single very high activity source of 192Ir is remotely inserted into the intracavitary applicator. Based on the treatment plan, during each treatment the source is advanced in individual “steps” to deliver radiation throughout the treatment volume. Because of the high activity of the source (usually about 10 Ci), treatment must be delivered in a heavily shielded room, and strict safety and quality assurance standards must be met. HDR therapy has become more popular in the past 15 years, particularly for intracavitary gynecologic applications. This is partly because of the practical advantages for physicians who center most of their practice in an outpatient clinical setting, but another factor is the recent interruption in the supply of cesium sources suitable for LDR gynecologic brachytherapy. Many clinicians remain reluctant to change to HDR therapy because of the theoretical radiobiological disadvantages of large-fraction irradiation and the absence of well-controlled randomized clinical trials comparing HDR and LDR regimens (69).
An alternative to HDR therapy that is commonly used in Europe but only recently introduced in the United States is pulsed dose-rate (PDR) brachytherapy. With this approach, treatment is given in intermittent pulses, using a single stepping source of 192Ir, similar to but lower in activity than the source used for HDR brachytherapy. If treatment is delivered in hourly pulses of 40 to 50 cGy, the tissue sparing should be nearly identical to that achieved with LDR brachytherapy. PDR holds several advantages over true LDR brachytherapy. The sources are readily obtainable, patients are able to receive nursing care and have visitors as they wish during the intervals between pulses, and the stepping source method permits somewhat more flexibility in treatment planning. The equipment can be used for either interstitial or intracavitary brachytherapy, and because the applicators are identical to those used for HDR, clinicians who choose to have both options available to their patients require only one set of applicators.
The total brachytherapy dose to point A must be reduced when converting from LDR to HDR regimens. The appropriate dose and dose per fraction is based on calculations of the estimated biologically effective dose (BED) on tumor and normal tissues. BED is derived from the linear-quadratic formula described earlier in this chapter and is equal to the total nominal dose (nd) times the relative effectiveness: BED = (nd) × (1 + d/α/β), where d is the dose per fraction. For example: Assuming α/β values of 10 and 3 for tumor and for normal tissues, respectively, a fractionation scheme in which a total dose of 30 Gy is given in five fractions of 6 Gy each would result in:
Tumor BED = (30) × (1 + 6/10) = 48 Gy10
Normal tissue BED = (30) × (1 + 6/3) = 90 Gy3
Clinicians often express these doses in the more familiar terms of the equivalent dose at 2 Gy per fraction, which is equal to BED/[1 + 2/(αβ)]. Using this calculation, the above example would yield equivalent doses of 40 and 54 Gy, respectively, for tumor and normal tissues. In other words, the effect on normal tissues is about 35% greater than would be expected from the same tumor-effective dose given at 2 Gy per fraction or with LDR brachytherapy (which, at 40 to 45 cGy per hour, has an effect similar to that of a dose divided in 2-Gy fractions).
Obviously, this differential effect would make HDR unacceptable if the normal tissues received the same dose as tumor. Fortunately, with good applicator positioning, effective packing of the bladder and rectum, and optimal source positioning, the total dose and dose per fraction delivered to normal tissues are usually considerably lower than those delivered to tumor, making it possible to achieve a ratio of tumor to normal tissue effect that is similar to what is achieved with LDR. If the tumor is very large or the vaginal anatomy is unfavorable, the nominal doses to tumor and normal tissues may be similar; in these cases, patients may be more effectively treated with LDR, PDR, or a larger than usual number of HDR fractions (69,70,71,72).
It is important that dose fractionation schemes used for HDR therapy produce tumor control and complication rates approximately equivalent to those seen with LDR therapy. The optimal dose per fraction of HDR therapy is unknown and is probably patient specific, but, in general, increasing the number of fractions and concomitantly decreasing the dose per fraction appears to reduce the rate of moderate and severe complications (71,73).
The most common HDR regimen used in the United States is probably five fractions of 5.5 to 6 Gy each to point A after 45 Gy to the pelvis, although there is wide variation in the number of fractions (2 to 13) and the dose per fraction (3 to 9 Gy) (74,75). Because large single fractions of radiation permit less recovery of sublethal injury than LDR irradiation, HDR therapy doses that yield a rate of tumor control equivalent to that seen with LDR therapy might result in an increased risk of late complications. However, with intracavitary treatment of the cervix, vulnerable normal tissues (primarily the rectum and bladder) are often some distance from the tumor site and therefore may receive a significantly lower dose and dose per fraction than the prescription point (usually point A).
Interstitial brachytherapy refers to the placement of radioactive sources within tissues. Various sources of radiation—such as 192Ir, 198Au, 103Pd, and 125I—may be obtained as radioactive wires or seeds. 192Ir may be obtained as separate sources that are usually distributed at regular intervals (usually 1 cm) in Teflon tubes or as wires with activity specified in terms of the mCi per cm. Sources may be positioned in the tumor or tumor bed in a variety of ways:
Most gynecologic interstitial implants are temporary LDR implants. Like intracavitary therapy, interstitial therapy delivers a relatively high dose of radiation to a small volume, sparing the surrounding normal tissues. However, the risk to normal tissues adjacent to the tumor or in the tumor bed may still be significant, particularly if the needle placement is inaccurate.
Figure 4.12 Interstitial implant for a stage II distal vaginal cancer. Needles are individually inserted transperineally; a finger is placed in the vagina while the needles are inserted to monitor the position of each needle relative to the tumor and mucosal surface. Left: A Lucite cylinder in the vagina displaces uninvolved vagina from the needles and has channels for additional sources at the periphery of the cylinder. Right: Postoperative radiographs show placement of the needles with a superimposed dose cloud that encloses the volume treated to 30 Gy.
Figure 4.13 Interstitial implant for an advanced cervical cancer. Reproduced from Dr. Mark Schray, Division of Radiation Oncology, Mayo Clinic, with permission.
Some investigators have advocated the use of template-guided interstitial brachytherapy to treat difficult cases of locally advanced cervical cancer (81,82) (Fig. 4.13). The ability to place sources in the lateral parametrium with this technique suggests a theoretical advantage over intracavitary treatment for patients with pelvic wall involvement. Some investigators have claimed high local control rates with this approach (81,82). However, survival rates are not clearly superior to those achieved with combined external-beam and intracavitary therapy, and the risk of major complications also may be greater (78,79).
The radiation oncology community remains polarized as to the appropriateness of interstitial therapy for patients with intact cervical carcinomas, and as yet no randomized trials have been conducted to compare the therapeutic ratio of conventional intracavitary irradiation with that of interstitial treatment. Interstitial implants may also be used in a variety of other gynecologic applications, including vaginal cancer, vaginal recurrence of cervical or endometrial cancer, and urethral cancer.
Intraperitoneal radioisotopes have been used to treat epithelial ovarian cancer in an effort to address the transperitoneal spread of the disease (83). Radioactive chromic phosphate (32P) has largely replaced colloidal gold (198Au) for peritoneal treatment. The longer half-life (14.3 days), pure β decay, and higher mean energy (0.698 MeV) of 32P yield slightly longer exposures, fewer radiation protection problems, and deeper tissue penetration than what is observed with 198Au.
If a radioisotope is evenly distributed within the peritoneum, it is theoretically possible to irradiate the entire peritoneal surface. However, the pattern of energy deposition within the abdomen and the dose delivered beneath the peritoneal surfaces depend on many factors, including the physical characteristics of the isotope used, the energies of its decay products, and the distribution of the isotope within the peritoneal cavity. In practice, isotope is seldom distributed uniformly to the peritoneal and omental surfaces (84). Postsurgical adhesions may limit the free flow of fluid, and this nonuniform distribution may result in underdosage of some peritoneal sites and overdosage of some normal tissues. This may result in unacceptable complications, particularly if intraperitoneal and external-beam irradiation are combined (85). Although randomized studies have demonstrated similar survival rates for patients with early ovarian cancer treated with 32P or single-agent chemotherapy, the role of intraperitoneal treatment still has not been clearly established (86), and this approach is rarely used today.
Clinical Uses of Radiation
Although specific radiation-therapy techniques may vary, the curative treatment of cervical cancer usually includes a combination of external pelvic irradiation and brachytherapy, often with concurrent chemotherapy. The goal of radiation therapy is to eliminate cancer in the cervix, paracervical tissues, and regional lymph nodes (64). All of these regions can be encompassed in a pelvic radiation field. However, the dose that can be delivered to the pelvis is limited by the tolerance of intrapelvic normal tissues, most importantly the rectosigmoid, bladder, and small bowel. Because the bulkiest tumor is usually in the cervix, this region typically requires higher doses than the rest of the pelvis to achieve locoregional control. Fortunately, it is usually possible to deliver these high doses with intracavitary therapy.
Typical external-beam fields are designed to include the primary tumor, paracervical tissues, and iliac and presacral lymph nodes, all with 1.5- to 2-cm margins. If the common iliac or aortic nodes are involved, then the treatment fields are usually extended to include at least the lower paraaortic region.
The borders of the typical anterior-posterior and posterior-anterior pelvic fields are as follows:
Every effort should be made to minimize the high-dose treatment volume while adequately encompassing the tumor and its regional lymph nodes. Using four beams (anterior, posterior, and right and left lateral) rather than an opposed pair of anterior and posterior beams (Fig. 4.8) can sometimes reduce the volume of tissue irradiated to a high dose.However, great care must be taken not to shield the primary tumor, uterosacral disease, or external iliac nodes when lateral fields are used (87,88). For some patients with locally advanced tumors, the amount of tissue spared with lateral fields may be relatively small after these areas are included. The additional bone marrow treated with lateral fields may also be a consideration if chemotherapy is part of the treatment plan. However, when the pelvis is treated after hysterectomy, four or more fields usually produce a more favorable dose distribution than two opposed fields. Some clinicians have advocated the use of highly conformal radiation-therapy techniques such as IMRT to treat the whole pelvis (89,90). When these highly conformal techniques are used, particular care must be taken to adequately cover the target volume and account for tumor response and internal organ motion.
For most patients with locally advanced disease, an initial course of treatment is given with external-beam irradiation and concurrent chemotherapy. Four to five weeks (40 to 45 Gy) of chemoradiation usually decreases endocervical disease and shrinks exophytic tumor, facilitating optimal intracavitary therapy. The dose to the central tumor is then supplemented with one or two LDR intracavitary treatments or with a variable number of HDR treatments. If the initial tumor volume is small or there is an excellent tumor response to external-beam irradiation and concurrent chemotherapy, then brachytherapy may be given earlier in the patient's treatment. Because the number of brachytherapy treatments is greater with HDR therapy than with LDR therapy, practitioners who use the HDR approach often begin brachytherapy before external-beam therapy has been completed, if the initial tumor response has been adequate. The balance between external-beam and intracavitary therapy may vary somewhat according to the tumor extent (64). However, several studies have suggested that intracavitary therapy is critically important to successful treatment, even for patients with very bulky stage IIIB tumors (1,91).
Patients with International Federation of Gynecology and Obstetrics (FIGO) stage IA disease can often be treated with intracavitary irradiation alone. Most patients with stage IB1 disease have a sufficiently high risk of metastasis to the pelvic lymph nodes to justify at least a moderate dose of pelvic radiation (e.g., 39.6 Gy) to sterilize possible microscopic regional disease.
For patients with carcinoma of the cervix who have vaginal hemorrhage, hemostasis can usually be achieved with vaginal packing, application of Monsel's solution, and rapid initiation of external-beam irradiation. For patients with excessive bleeding, transvaginal irradiation (if available) or several days of accelerated pelvic radiation therapy (e.g., 1.8 Gy twice daily) may be helpful.
The total doses of radiation to the central tumor and regional nodes are tailored according to the amount of disease in those sites (92). Anumber of methods have been used to prescribe and specify the doses delivered with intracavitary therapy. Most radiation oncologists specify treatment using some variation of the Manchester system, which uses two primary reference points (Fig. 4.11):
Although the cGy doses from intracavitary and external-beam radiation therapy may not be biologically equivalent (particularly with HDR therapy), these doses are frequently summed to determine the total doses to points A and B. The total dose to point A (from external-beam and LDR intracavitary therapy) believed to be adequate to achieve central disease control is usually between 75 Gy (for small stage IB1 cancers) and 90 Gy (for bulky or locally advanced disease). The prescribed dose to point B is 45 to 65 Gy, depending on the extent of parametrial and sidewall disease.
Prescription and treatment planning cannot be limited to specification of the dose to these reference points. Other factors that should be considered include the following:
A number of methods and reference doses have been described to estimate the maximum dose to the bladder and rectum on the basis of orthogonal reference films of the implants. The most common method for specifying normal tissue doses is to calculate the doses to reference points defined by the International Commission on Radiation Units and Measurements (Figure 4.11) (93). Using this method, the bladder reference point is placed at the posterior edge of a Foley bulb filled with 7 cc of contrast material; the rectal point is located 5 mm posterior to the vaginal applicator or packing (whichever is most posterior) at the level of the vaginal sources. Three-dimensional reconstructions of intracavitary placements suggest that most methods that use orthogonal x-rays to estimate the dose to normal structures tend to underestimate the true maximum dose (94). For this reason, it is important to qualitatively examine each intracavitary system rather than to depend solely on normal-tissue reference points.
Some centers also document the total milligram-Radium-equivalent hours (mgRaEq-hr) of each intracavitary system. This number, obtained by multiplying the mgRaEq of cesium or radium in the system by the number of hours the radioactive sources are left in place, cannot be used as the sole measure of any treatment but is sometimes used to limit the total integral dose to the pelvis. The doses to points at a substantial distance from the system are roughly proportional to the total mgRaEq-hr because, as the distance increases, the dose rate approaches that from a single point source of similar activity. In general, after 40 to 45 Gy of external-beam irradiation, the total mgRaEq-hr (from intracavitary radiation therapy given at 40 to 60 cGy/hr) should not exceed 6,000 to 6,500. An alternative measure is the reference air kerma, defined as the total dose delivered at 1 meter from the center of the activity and measured in µy.m2; this unit serves the same purpose but can more easily be used with isotopes other than radium or cesium.
There is a growing movement toward use of image-guided brachytherapy (IGBT) with treatment planning based on CT or MRI images obtained with the implant in place. Ideally, IGBT would include true three-dimensional imaging and calculations of the distribution of dose to tumor and normal tissues with accompanying dose limits derived from clinically derived dose-effect data. There currently are very few clinical data derived from patients treated with IGBT; logistical problems also create significant impediments to true IGBT. As a first step, many clinicians are obtaining at least ultrasonic or CT images of intracavitary systems to rule out unsuspected uterine perforation (a potential source of serious complications) and to provide closer estimates of relationships between critical structures and radiation sources.
Results of Treatment
Radiation therapy is extremely effective in the treatment of stage IB1 cervical cancer, producing central and pelvic disease control rates of greater than 98% and greater than 95%, respectively, and disease-specific survival rates of approximately 90% (2,4). Pelvic control rates decrease as tumor size and FIGO stage increase, although large singleinstitution experiences report 5-year pelvic control rates of 60% to 70% and disease-specific survival rates of 40% to 50% even for bulky stage IIIB cancers treated with radiation alone (e.g., before routine use of concurrent chemotherapy) (1,2). Although these control rates clearly indicated a need for more effective treatment of these advanced lesions, it is remarkable that such massive carcinomas, usually more than 7 cm in diameter, can be controlled even half the time with radiation therapy alone. This undoubtedly reflects the remarkable effectiveness of carefully planned combinations of external-beam and intracavitary radiation therapy.
During the past decade, studies have demonstrated a significant improvement in pelvic disease control and survival when cisplatin-containing chemotherapy is delivered concurrently with radiation for patients with locoregionally advanced cervical cancer (19,20,21,95,96,97). Several of the regimens tested in these studies also included 5-fluorouracil. This drug is known to be a potent radiation sensitizer, particularly effective in the treatment of gastrointestinal malignancies, but its contribution to chemoradiation in patients with carcinoma of the cervix is still uncertain. Other randomized trials suggest that mitomycin C (98, 99) and epirubicin (100) also may improve outcome when they are delivered concurrently with radiation to patients with locally advanced cervical cancer.
Adjuvant Pelvic Radiation Therapy after Radical Hysterectomy
For patients with stage IB and IIA cervical cancer treated with radical hysterectomy and pelvic lymphadenectomy, lymph node involvement is probably the strongest predictor of local recurrence and death: Patients with nodal involvement have survival rates of only 50% to 60% of those of patients with negative nodes (101,102). Parametrial involvement and involvement of surgical margins also predict a high rate of pelvic recurrence and are considered to be indications for postoperative irradiation. In 2000, the Southwest Oncology Group published results of a study comparing postoperative radiation with combined chemoradiation in patients who had positive lymph nodes, parametrium, or surgical margins; the study demonstrated a 50% reduction in the risk of recurrence when cisplatin andfluorouracil were added to pelvic irradiation (96).
For patients with negative nodes but high-risk primary tumor features (i.e., tumor size ?4 cm, deep stromal invasion, or vascular space involvement), postoperative irradiation has also been demonstrated to produce a significant reduction in the risk of recurrence (7,103). This is discussed further in Chapter 9.
The drawback of adjuvant pelvic radiation therapy is a somewhat greater risk of major complications than with surgery alone or radiation alone (5,7). For this reason, a National Cancer Institute Consensus Conference (60) concluded that “primary therapy should avoid the routine use of both radical surgery and radiation therapy,” suggesting that patients who are known to have high-risk factors at initial evaluation may be better treated with radical radiation therapy.
Recurrent Cervical Cancer
Patients who have an isolated pelvic recurrence after radical hysterectomy can sometimes Cancer be treated successfully with aggressive radiation therapy. The prognosis is best for patients with isolated central recurrences that are not fixed to the pelvic wall and do not involve pelvic nodes. These patients have 5-year survival rates as high as 60% to 70% (104). For patients whose tumors involve the pelvic wall or lymphnodes; a few groups report better than a 20% 5-year survival rate for radiation alone. Some groups have reported encouraging results with combined chemoradiation (105). It probably is reasonable to extrapolate from recent randomized trials that demonstrate improved survival with concurrent chemoradiation for locally advanced cervical cancer to justify a similar approach in patients with pelvic recurrences.
Late complications of radical irradiation for cervical cancer occur in 5% to 15% of patients and are related to the dose per fraction, the total dose administered, and the volume irradiated (106). Patient factors such as a history of pelvic infection, heavy smoking, previous abdominal surgery, and diabetes mellitus may increase the risk of complications (56,107). The positioning of the intracavitary system also may influence the risk of complications. Late effects may be seen in the bladder (hematuria, fibrosis and contraction, or fistulas) and in the rectosigmoid or terminal ileum (bleeding, stricture, obstruction, or perforation). Agglutination of the apex of the vagina is common. Severe vaginal shortening is less common and is probably correlated with the patient's age, menopausal status, and sexual activity and with the initial extent of disease (108,109). Unfortunately, our understanding of the factors influencing sexual dysfunction in patients treated for cervical cancer is still incomplete. Most gastrointestinal complications occur within 30 months of radiation therapy, although late effects may occur many years after treatment (109).
In the United States, late complications of radiation therapy (occurring more than 90 days after treatment) are usually scored according to the Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancer Late Radiation Morbidity Scheme, which is part of the National Cancer Institute system for reporting of adverse events (Table 4.2). However, it is important to recognize that with today's multimodality treatments, several factors may contribute to adverse events. In Europe, many groups use the Franco-Italian Glossary, a scoring system that incorporates early and late surgical and radiation-related side effects (110).
Radiation therapy plays an important role in the palliation of metastatic cervical cancer. Short courses of palliative irradiation, such as 2,000 cGy in five fractions or 3,000 cGy in 10 fractions, usually will alleviate symptoms related to bony metastases or paraaortic nodal disease. Such treatment also may relieve symptoms related to pressure from enlarging mediastinal or supraclavicular nodal disease.
Table 4.2 RTOG and EORTC Late Radiation Morbidity Scoring Schemea
The role of radiation therapy in the treatment of endometrial carcinoma is discussed in greater detail in Chapter 10. Indications for radiation therapy in the treatment of endometrial cancer are as follows:
In the past, disease confined to the uterus was often treated with preoperative intracavitary radiation therapy. An intracavitary line source was placed in the uterus, or the uterus was packed with multiple radium (Heyman's) capsules or cesium (Simon's) capsules (111). Preoperative irradiation reduces the risk of vaginal apex recurrence but has never been proven to improve survival, although no randomized studies have been done to compare preoperative and postoperative irradiation (112,113). Because tailored postoperative irradiation appears to achieve similar pelvic control rates and avoids unnecessary over-treatment of patients whose hysterectomy findings predict a negligible risk of recurrence, preoperative irradiation has been abandoned for most patients (112,114,115).
Most patients with stage I endometrial cancer have minimally invasive grade 1-2 tumors, which rarely recur after hysterectomy alone and usually need no additional treatment. The use of adjuvant pelvic radiation therapy is usually confined to patients with deeply invasive lesions or other high-risk findings at surgery (e.g., lymph node involvement or cervical stromal involvement) (112,114,115). Adjuvant pelvic radiation therapy reduces the risk of pelvic recurrence but has never been proven to improve survival. In 2004, the Gynecologic Oncology Group reported results of a randomized trial addressing this question in patients with intermediate-risk FIGO stage I cancers. This study demonstrated a reduction in the overall risk of pelvic (particularly vaginal) recurrence without a significant difference in overall survival for patients who received postoperative pelvic radiation therapy (116). However, using subset analysis, the authors identified a subset of patients with high-intermediate-risk disease who may benefit from adjuvant radiation therapy. In another randomized trial, Creutzberg et al. (58) found that postoperative radiation therapy reduced the risk of pelvic recurrence but had no significant impact on survival. Unfortunately, both of these trials included a large number of patients who had relatively favorable findings (grade 1 disease or <50% invasion); neither trial included a sufficient number of patients who had grade 3 tumors or deep myometrial invasion to rule out clinically important differences in these subgroups.
Uterine papillary serous cancers have a particularly poor prognosis and an inclination to spread intraperitoneally in a manner similar to that seen with ovarian cancers. Whole-abdominal irradiation appears to be valuable treatment for some patients with minimal residual disease after hysterectomy (117,118), although many groups favor the use of adjuvant chemotherapy with or without local radiation therapy for this group of patients.
For patients with endometrial cancer, the potential benefit of adjuvant treatment must be balanced against the risk of complications for each patient. Extensive staging lymphadenectomy appears to increase the risk of serious bowel complications after radiation therapy (119,120).
Several independent investigators have established a curative role for whole-abdominal and pelvic irradiation for some subsets of patients with epithelial ovarian cancer (121,122,123,124). Survival rates are strongly correlated with initial disease stage and volume of residual disease. The best survival rates are for patients with stage II disease and for those whose macroscopic residual disease was confined to the pelvis—a situation in which a relatively high dose of radiation can be given.
Because transperitoneal spread is the most common route of dissemination of ovarian cancer, radiation fields that encompass the whole peritoneal cavity are more likely to be curative than those that treat only the pelvis or lower abdomen. However, normal tissues in the upper abdomen (e.g., kidney, liver, bowel, and spinal cord) limit the dose of radiation that can be delivered to the whole abdomen to approximately 22 Gy. Somewhat higher doses can be delivered to portions of the upper abdomen that do not include the most sensitive normal structures. Because 22 Gy is insufficient to control macroscopic disease, patients with gross upper abdominal disease cannot be expected to benefit from whole-abdominal irradiation. Although a curative benefit has been established for whole-abdominal radiation therapy, randomized studies have never determined the relative benefits of abdominopelvic radiation therapy and combination platinum-based chemotherapy in appropriately selected patients with minimal residual disease.
Radiation therapy may also be helpful in the palliation of pain, bleeding, or other symptoms from ovarian cancer. Occasionally, patients who have localized recurrences may experience prolonged disease-free survival after localized radical radiation therapy, particularly if all macroscopic disease can be surgically resected initially.
The high response rates but frequent relapses observed after treatment of ovarian cancer with chemotherapy have encouraged investigators to add whole-abdominal irradiation either as a salvage treatment for incomplete responses or as part of an up-front multimodality treatment program. Many small single-arm studies of sequential treatments have been reported, and they show varied results. A retrospective analysis of the Toronto data suggested an improved outcome for high-risk patients treated with sequential chemotherapy and whole-abdominal irradiation compared with historical controls treated with radiation alone (125). However, three randomized studies have compared chemotherapy alone with multimodality treatment (126,127,128) with disappointing results. Although some patients with minimal residual disease may benefit, in general the data do not support routine use of sequential chemotherapy and abdominopelvic irradiation. Poor tolerance after extensive chemotherapy and the possible induction of accelerated repopulation of resistant clonogens during treatment are among the reasons suggested for the failure of this approach in most hands (121,129,130).
Early studies of whole-abdominal radiation therapy used a “moving-strip” technique in which a 10-cm-high field (usually 60Co) was moved by 2.5-cm increments so that each “strip” received 8 or 10 fractions, usually of 2.25 Gy each (131,132). With the advent of high-energy linear accelerators, this approach was replaced by an open-field technique (Fig. 4.14). These two techniques have been compared in randomized trials that demonstrated no significant difference in survival and a lower rate of bowel complications (1% versus 6%) with the openfield technique (132,133). Most abdominopelvic irradiation techniques include a boost to the pelvis, and some investigators boost the paraaortic nodes and medial diaphragms (“T boost”) to 40 to 45 Gy after initial whole-abdominal treatment (123). The design of abdominopelvic fields requires careful simulation using fluoroscopy and CT-based planning to confirm adequate coverage of the peritoneal surfaces and diaphragms and proper shielding of sensitive structures.
Figure 4.14 Treatment portals for carcinoma of the ovary or for uterine papillary serous carcinoma. The field must encompass the entire peritoneal cavity. Shielding is usually added to limit the dose to the kidneys to less than 18 to 20 Gy. The liver dose is usually limited to 25 Gy.
Acute side effects of abdominopelvic irradiation include nausea, anorexia, general fatigue, and diarrhea in most patients (134). These symptoms are usually fairly well controlled with appropriate medications. Approximately 10% of patients develop significant myelotoxicity (platelet count <100,000 or neutrophil count <1,500). The risk of significant toxicity is much higher in patients who undergo abdominopelvic irradiation after chemotherapy, and the risk depends on the drugs and the duration of previous chemotherapy. As many as 40% of patients treated with abdominopelvic irradiation have transiently elevated levels of alkaline phosphatase, but symptomatic hepatitis is rare if the dose of radiation to the liver does not exceed 27 Gy. In the absence of tumor recurrence, late bowel complications are rare, but the risk tends to increase with the extent and number of previous abdominal operations (particularly lymphadenectomy) (135).
The role of radiation therapy in the treatment of vulvar cancer has increased dramatically during the past 25 years. Improved radiation-therapy equipment and techniques have reduced the toxicity that discouraged early attempts to treat the vulva with radiation, and prospective studies have increased interest in this effective modality. In particular, the landmark randomized study published by Homesley and colleagues in 1986 demonstrated a marked improvement in survival when patients with positive lymph nodes were treated with pelvic and inguinal irradiation after vulvectomy and lymphadenectomy (16). The role of radiation for treatment of vulvar cancer is explored in more detail in Chapter 13.
In brief, the possible benefits of radiation therapy in the treatment of vulvar cancer include: (i) reduced risk of regional recurrence and improved survival in patients with inguinal node metastases (16); (ii) reduced risk of vulvar recurrence in patients with positive surgical margins, multiple local recurrences, or other high-risk features (136,137); and (iii)avoidance of exenterative surgery in patients whose disease involves the anus or urethra (138). Radiation therapy may also be an alternative to inguinal lymphadenectomy in selected patients with clinically negative groins (17,139).
Several reports have emphasized the critical importance of careful radiation-therapy technique in treating patients with vulvar cancer (17,139,140). A number of approaches have been developed to decrease the dose to the femoral heads from groin irradiation. In most cases, adequate coverage of the volume at risk is readily achieved without risking serious femoral morbidity. However, this can only be accomplished with detailed CT-based treatment planning.
In general, the total dose of radiation should be tailored to the amount of residual disease, with doses of approximately 45 to 50 Gy for microscopic disease and 60 Gy or higher for positive margins, extracapsular nodal extension, or macroscopic residual disease.
When necessary, the dose to portions of the vulva at high risk for recurrence can be “boosted” with an en face electron field. This approach minimizes the amount of tissue exposed to high doses and thereby reduces acute skin reactions. Carefully designed IMRT may be useful in selected cases. Bolus may be needed to increase the dose to superficial tissues in the “buildup region” of photon and low-energy electron beams. Treatment interruptions should be minimized to avoid possible tumor proliferation during breaks in radiation therapy.
The use of concurrent “sensitizing” chemotherapy (e.g., continuous-infusion fluorouracil or cisplatin) to improve control rates has been explored in a number of uncontrolled studies (141,142,143,144,145,146,147). The encouraging response rates and long-term control of gross disease reported in these trials and the successful use of chemoradiation in cervical and anal cancer are bound to increase interest in this approach in the future.
Acute moist desquamation of the skin of the inguinal creases and vulva is expected. Symptoms may be reduced with careful local care, sitz baths, avoidance of tight clothing, and immediate treatment of superimposed fungal or bacterial infections. Superinfection with Candida species is particularly frequent during treatment. Late complications may include lymphedema, particularly after radical groin dissection. Atrophy, telangiectasia, and fibrosis of the skin or subcutaneous tissues can occur and may be related to the daily fraction size and total dose, tissue destruction from tumor, and the extent of local surgery.
Although small apical vaginal lesions can sometimes be resected, the intimate relationship of the vagina to the bladder and rectum usually makes it impossible to perform curative surgical resection without sacrificing those organs. For this reason, most patients who have invasive vaginal cancers are treated with radiation therapy, which achieves cure rates that are, stage for stage, similar to those achieved with radiation therapy for patients with cervical cancer (13,14,148). Treatment usually consists of a combination of external-beam irradiation and brachytherapy. Interstitial or intracavitary techniques may be used, depending on the size and site of the primary lesion and its response to external-beam therapy (Fig. 4.12). Because the dose gradient from intracavitary therapy is very steep, interstitial techniques are usually used to treat tumors that are more than 3 to 5 mm thick. Tumors that are very advanced, diffuse, or fixed or that extensively involve the rectovaginal septum may be boosted to a high dose using conformal external-beam therapy or IMRT. The vaginal apex can move 2 to 3 cm with bladder filling and emptying; this internal organ motion should be carefully considered during treatment planning.
Concurrent chemoradiation may have a role in the treatment of vaginal cancers, although there are no large trials in this group of patients. Many clinicians believe that similarities in histology and behavior justify the use of regimens that have been proven beneficial for cervical cancer to treat locally advanced vaginal cancers. Also, because vaginal cancer is very rare and the radiation-therapy techniques are specialized, patients with this disease may benefit from referral to centers with relatively large gynecologic radiation oncology practices.