Berek and Hacker's Gynecologic Oncology, 5th Edition
Tumor Growth and Chemotherapy
A wide variety of chemotherapeutic agents is available, and the selection and dose of drugs is determined by the relative benefits of individual agents, as well as combinations, based on the results of phase II and phase III trials. Most antineoplastic agents have a narrow therapeutic index, and the choice of treatment is based on several factors, including age, performance status, comorbidities, organ function, tumor type, and whether or not a patient has had previous chemotherapy (Table 3.1).
Table 3.1 Issues to Be Considered before Using Antineoplastic Drugs
It is important to have a good understanding of the likely natural history of each patient's malignancy. This may influence decisions regarding when to initiate treatment, particularly when chemotherapy is being administered with palliative intent, and it also may be appropriate to withhold chemotherapy in asymptomatic patients with indolent metastatic disease. Chemotherapy should be restricted to patients in whom the diagnosis of cancer has been confirmed by either biopsy or cytology.
All chemotherapeutic agents have potential side effects, and it is important to ascertain whether the patient has measurable disease or elevated tumor markers before commencing treatment, particularly in patients with metastatic disease, so that response can be assessed objectively. The extent of previous therapy and the patient's age, general health, and other relevant medical problems (e.g., neuropathy from long-standing diabetes) are all taken into consideration when deciding on the choice of treatment. In addition, the patient's emotional, social, and financial status must be respected and taken into consideration. It is also important to clearly communicate the aims and objectives of therapy, the likelihood of benefit, and the side effects of treatment so the patient can make an informed decision regarding chemotherapy.
Tumors can be grouped into the following four categories by their likelihood of chemotherapeutic response and benefit from treatment:
- Highly chemosensitive tumors—treatment administered with curative intent. This includes ovarian germ cell tumors and gestational trophoblastic tumors where chemotherapy is curative for most patients. Toxicity is acceptable if the probability of cure is high, but every effort must be made to limit toxicity and reduce side effects without compromising the chance of cure such as by inappropriate dose reductions or prolonged delay between cycles.
- Chemosensitive, but cure is uncommon. This includes advanced epithelial ovarian cancer, where response rates are in excess of 70% to 89%, but most patients will eventually relapse. Chemotherapy improves survival but does not restore a normal life expectancy in the majority of patients. Individuals with these tumors usually benefit from chemotherapy, and it should be offered unless there are exceptional circumstances.
- In the third group (e.g., uterine leiomyosarcoma), responses to chemotherapy are relatively low, and the impact of treatment on overall survival is unclear. In this setting, it is not unreasonable to consider a patient's overall medical condition in the determination of the relative risks versus benefits of cytotoxic drug therapy.
- Relatively chemoresistant tumors (e.g., metastatic melanoma) where responses to chemotherapy are low and very unpredictable. If treatment is considered, every effort should be made to include these patients in well-designed, prospective clinical trials testing new treatment approaches.
Chemotherapeutic agents must have greater activity against tumors than against normal tissues. The therapeutic window between antitumor effect and normal tissue toxicity may be small because most chemotherapeutic agents work by disrupting DNA or RNA synthesis, affecting crucial cellular enzymes, or altering protein synthesis.
Normal cells also use these vital cellular processes in ways similar to those of malignant cells, particularly fetal or regenerating tissue or normal cell populations in which constant cell proliferation is required (e.g., bone marrow, gastrointestinal epithelium, and hair follicles). As a result, the differential effect of antineoplastic drugs on tumors compared with normal tissues is quantitative rather than qualitative, and some degree of injury to normal tissue is produced by every chemotherapeutic agent. The normal tissue toxicity produced by most chemotherapeutic agents correlates with the intrinsic cellular proliferation of the target tissue. Hence, blood count suppression, mucosal injury, and alopecia are commonly seen with most chemotherapeutic regimens.
For any particular chemotherapeutic agent, the net effect on the patient is often referred to as the drug's therapeutic index (i.e., a ratio of the doses at which therapeutic effect and toxicity occur). Cancer chemotherapy requires a balance of therapeutic effect and toxicity to optimize the therapeutic index. Because the window of toxicity is often narrow for available chemotherapeutic agents, successful chemotherapy depends on pharmacologic and biologic factors.
Biologic Factors Influencing Treatment
Cell Kinetic Concepts
Both normal and tumorous tissues have a certain growth capacity and are influenced and regulated by various internal and external forces. The differential growth and regulatory influences occurring in both normal and tumorous tissues form the basis of effective cancer treatment. The exploitation of these differences forms the basis for the effective use of both radiation and chemotherapy in cancer management.
Patterns of Normal Growth
All normal tissues have the capacity for cellular division and growth. However, normal tissues grow in substantially different patterns. There are three general types of normal tissue growth: static, expanding, and renewing.
- The static population comprises relatively well-differentiated cells that, after initial proliferative activity in the embryonic and neonatal period, rarely undergo cell division. Typical examples are striated muscle and neurons.
- The expanding population of cells is characterized by the capacity to proliferate under special stimuli (e.g., tissue injury). Under those circumstances, the normally quiescent tissue (e.g., liver or kidney) undergoes a surge of proliferation with regrowth.
- The renewing population of cells is constantly in a proliferative state. There is constant cell division, a high degree of cell turnover, and constant cell loss. This occurs in bone marrow, epidermis, and gastrointestinal mucosa.
Normal tissues with a static pattern of growth are rarely seriously injured by drug therapy, whereas renewing cell populations such as bone marrow, gastrointestinal mucosa, and spermatozoa are commonly injured.
Cancer Cell Growth
Tumor cell growth represents a disruption in the normal cellular brake mechanisms, resulting in continued proliferation and eventual death of the host. It is not the speed of cell proliferation but the failure of the regulated balance between cell loss and cell proliferation that differentiates malignant cells from normal cells.
The characteristics of cancer growth have been assessed by multiple studies in animals and more limited studies in humans. When tumors are extremely small, growth follows an exponential pattern but later seems to slow. Such a growth pattern is known as Gompertzian growth. Strictly speaking, this means exponential growth with exponential growth retardation over the entire duration of tumor growth. More simply, Gompertzian growth means that, as a tumor mass increases in size, the time required to double the tumor's volume also increases.
The doubling time of a human tumor is the time it takes for the mass to double its size. There is considerable variation in doubling times of human tumors. For example, embryonal tumors, lymphomas, and some malignant mesenchymal tumors have relatively fast doubling times (20 to 40 days), whereas adenocarcinomas and squamous cell carcinomas have relatively slow doubling times (50 to 150 days). In general, metastases have faster doubling times than primary tumors.
If it is assumed that exponential growth occurs early in a tumor's history and that a tumor starts from a single malignant stem cell, then
- a 1-mm mass will have undergone approximately 20 tumor doublings
- a 5-mm mass (a size that might be first visualized on a radiograph) will have undergone 27 doublings
- a 1-cm mass will have undergone 30 doublings
Were such a lesion discovered clinically, the physician would assume that the tumor had been detected early. The reality is that it would have already undergone 30 doublings or been present approximately 60% of its life span.
Unfortunately, our current diagnostic methods detect tumors only relatively late in their growth, and metastases may well have occurred long before there is obvious evidence of the primary lesion. The second implication of tumor kinetics is that, in late stages of tumor growth, a very few doublings in tumor mass have a dramatic impact on the size of the tumor. Once a tumor becomes palpable (1 cm in diameter), only three more doublings would produce an enormous tumor mass (8 cm in diameter).
Information on growth patterns and doubling times relates to the growth of the tumor mass as a whole. The kinetic behavior of individual tumor cells has been well described, and a classic cell cycle model has been produced (Fig. 3.1).
- M phase (mitotic phase) of the cell cycle is the phase of cell division.
- G1phase (postmitotic phase) is a period of variable duration when cellular activities and protein and RNA synthesis continue. These G1 cells can differentiate or continue in the proliferative cycle.
- S phase (DNA synthetic phase) is the period in which new DNA replication occurs.
- G2phase (postsynthetic phase) is the period in which the cell has a diploid number of chromosomes and twice the DNA content of the normal cell. The cell remains in this phase for a relatively short time and then enters the mitotic phase again.
- G0phase (the resting phase) is the time during which cells do not divide. Cells may move in and out of the G0 phase.
The generation time is the duration of the cycle from M phase to M phase. Variation occurs in all phases of the cell cycle, but the variation is greatest during the G1 period. The reasons for this variation are complex and not completely understood.
Figure 3.1 The cell cycle. After cell division, a cell can (1) die, (2) differentiate, or (3) enter resting (G0) phase. Cells in the latter two phases can reenter the cycle at G1.
These cell cycle events have important implications for the cancer therapist. Differential sensitivities to chemotherapy and radiation therapy are associated with different proliferative states. Dividing cancer cells that are actively traversing the cell cycle are very sensitive to chemotherapeutic agents. Cells in a resting state (G0) are relatively insensitive to chemotherapeutic agents, although they occupy space and contribute to the bulk of the tumor.
In cell kinetic studies performed on human tumors, the duration of the S phase (DNA synthesis phase) is relatively similar for most human tumors, ranging from a low of 10 hours to a high of approximately 31 hours. The length of the cell cycle in human tumors varies from slightly more than half a day to perhaps 5 days. With cell cycle times in the range of 24 hours and doubling times in the range of 10 to 1,000 days, it is clear that only a small proportion of tumor cells are in active cell division at any one time.
Two major factors that affect the rate at which tumors grow are the growth fraction and cell death. The growth fraction is the number of cells in the tumor mass that are actively undergoing cell division. There is a marked variation in the growth fraction of tumors in human beings, ranging from 25% to almost 95%. In the past, it was thought that human tumors contained billions of cells, all growing slowly. In actuality, only a small fraction of cells in a tumor mass are rapidly proliferating; the remainder are out of the cell cycle and quiescent. Cancer “stem cells” are a very small population of cells that appear to be relatively chemoresistant; these play a major role in the development and progression of cancers.
Tumor growth may be altered by the following:
- cytotoxic chemotherapy, which alters both the generation time and the growth fraction of a tumor
- hormones, which appear to alter the growth fraction without changing the generation time
- radiation therapy, which alters both the generation time and the growth fraction
- alterations in oxygen tension and vascular supply, which alter the growth fraction without altering generation time
- immunologic therapies, which seem to alter both generation time and growth fraction
Cell Cycle-Specific versus Cell Cycle-Nonspecific Drugs
Antineoplastic agents have complex mechanisms of action and alter cells in a wide variety of ways. Different drugs have different sites of action in the cell cycle, and their effectiveness is also a function of the proliferative capacity of the tissue involved. With the use of some of these kinetic concepts, it is possible to classify chemotherapeutic agents on the basis of their cell cycle specificity and their site of maximal drug action within the cell cycle (Table 3.2).
Cell Cycle-Nonspecific Cell cycle-nonspecific agents kill in all phases of the cell cycle and are not too dependent on proliferative activity.
Cell Cycle-Specific Cell cycle-specific agents, such as hydroxyurea, depend on the proliferative activity and on the phase of the cell cycle for their action. The agents kill in only one phase of the cell cycle, and cells not in that phase are not injured. They tend to be most effective against tumors with relatively long S phases and against those tumors in which there is a relatively high growth fraction and a rapid rate of proliferation. Between these two broad classifications, there is a spectrum of drugs with variable degrees of cell cycle and proliferation dependence.
Table 3.2 Cell Cycle Specificity of Chemotherapeutic Agents
Table 3.3 Site of Action in the Cell Cycle
In addition to cell cycle and proliferative sensitivity, chemotherapeutic agents may exert a greater effect in a particular phase of the cell cycle. Thus, chemotherapeutic agents can be grouped according to their site of action in the cell cycle and the extent of their dependence on proliferative activity (Table 3.3).
Log Kill Hypothesis
From knowledge of basic cellular kinetics, there have emerged certain concepts of chemotherapy that have proven useful in the design of chemotherapeutic regimens. In experimental tumor systems in animals, survival of the animal is inversely proportional to the number of cells implanted or to the size of the tumor at the time treatment is initiated (1). Treatment immediately after tumor implantation or when the tumor is subclinical in size results in more cures than when the tumor is clinically obvious and large.
Chemotherapeutic agents appear to work by first-order kinetics; that is, they kill a constant fraction of cells rather than a constant number. This concept has important conceptual implications in cancer treatment. For instance, a single exposure of tumor cells to an antineoplastic drug might be capable of producing 2- to 5-logs of cell kill. With typical body tumor burdens of 1012 cells (1 kg), a single dose of chemotherapy is unlikely to be curative. This explains the need for intermittent courses of chemotherapy to achieve the magnitude of cell kill necessary to produce tumor regression and cure. It also provides a rationale for multiple-drug or combination chemotherapy.
The cure rate would be significantly improved if small tumors were present, but cell masses of 101 to 104 cells are too small for clinical detection. This is the basis for using adjuvant chemotherapy in early stages of disease when subclinical metastases are likely to be present in many patients.
Drug Resistance and Tumor Cell Heterogeneity
Chemotherapeutic agents often are active when initially used in cancer treatment, but tumors commonly become resistant during chemotherapy. Hence, patients often have an initial remission followed by a recurrence that is no longer responsive to the drugs that were initially effective.
Various cellular mechanisms are involved in drug resistance. Resistant tumor cells may display increased deactivation or decreased activation of drugs, allow increased drug efflux, or resist normal drug uptake. In some instances, altered specificity to an inhibiting enzyme or increased production of the target enzyme occurs to explain drug resistance on a pharmacologic basis.
Theories for Overcoming Drug Resistance
It has been suggested that spontaneous mutation to phenotypic drug resistance occurs in rapidly growing malignant tumors: This is the somatic mutation theory (2). The theory suggests that most mammalian cells start with intrinsic sensitivity to antineoplastic drugs but develop spontaneous resistance at variable rates. This concept—the Goldie-Coldman hypothesis—has been applied to the growth of malignant tumors and has important clinical implications.
Goldie and Coldman developed a mathematical model that relates curability to the time of appearance of singly or doubly resistant cells. Assuming a natural mutation rate, the model predicts a variation in size of the resistant fraction in tumors of the same size and type, depending on the mutation rate and the point at which the first mutation develops. Given such assumptions, the proportion of resistant cells in any untreated tumor is likely to be small, and the initial response to treatment would not be influenced by the number of resistant cells. In clinical practice, this means that a complete remission could be obtained even if resistant cells were present. The failure to cure such a patient, however, would be directly dependent on the presence of resistant cells.
This model of spontaneous drug resistance implies that:
- Tumors are curable with chemotherapy if no permanently resistant cells are present and if chemotherapy is begun before resistance develops.
- If only one antineoplastic agent is used, then the probability of cure diminishes rapidly with the development of a single resistant line.
- Minimizing the emergence of drug-resistant clones requires multiple effective drugs and that they be applied as early as possible in the course of the patient's disease.
- The rate of spontaneous mutation to resistance occurs at approximately the natural frequency of 1 in 10,000 to 1 in 1,000,000 cell divisions.
This model predicts that alternating cycles of treatment should be superior to the sequential use of particular agents because sequential use of antineoplastic drugs would allow for the development and regrowth of a doubly resistant line. The intrinsic frequency of spontaneous mutation to drug resistance is also likely to be influenced by etiological factors responsible for tumor development. Lung or bladder cancers, for instance, result from exposure to multiple carcinogenic chemicals and may have a higher spontaneous mutation rate than is seen in other tumors. Under these circumstances, numerous drug-resistant clones may be present even before the tumors are clinically evident. This would explain the inability of antineoplastic therapy to cure a number of the common malignancies.
An alternative hypothesis, developed by Norton and Simon, focuses on the Gompertzian growth rates exhibited by malignant tumors (3,4). This mathematical model suggests that the efficacy of treatment of tumors exhibiting sensitivity to particular chemotherapeutic agents will be enhanced if single agents, or combination regimens, are delivered at their optimal dose levels in a so-called dose-dense manner rather than as alternating regimens.
The fundamental difference between the Norton-Simon and Goldie-Coldman models is that in the former approach, the individual drugs are given in sequence at their optimal levels to produce a cytotoxic effect, whereas in the later strategy, which focuses on the rapid administration of as many active agents as possible, dose levels of individual drugs will frequently need to be modified because of overlapping toxic effects (e.g., bone marrow suppression).
Randomized trials in breast cancer have provided important evidence in support of the Norton-Simon hypothesis, with novel strategies being designed to deliver active drugs in the dosedense manner. High-risk gestational trophoblastic tumors are very chemosensitive, and treatment with EMA-CO every 6 to 7 days is an example of a dose-dense regimen.
Pleiotropic Drug Resistance
If the failure of drug treatment depends on the spontaneous appearance of resistant cells, an understanding of drug resistance is crucial to therapeutic success. A wide variety of mechanisms for drug resistance has been described, although these mechanisms usually confer resistance to a particular drug or drug family. The phenomenon of pleiotropic drug resistance occurs when certain drug-resistance mechanisms confer crossresistance to structurally dissimilar drugs with different mechanisms of action (5).
Some pleiotropic resistant cells contain a cell surface P glycoprotein with a molecular weight of 170 kilodaltons. In general, the appearance of pleiotropic drug resistance is associated with the cell's impaired ability to accumulate and retain antineoplastic drugs. It has been further demonstrated that this P glycoprotein is directly related to the expression of resistance, and cells that revert to sensitive ones lose this membrane glycoprotein.
Dose Intensity and High-Dose Chemotherapy
Studies in human solid tumors in vitro frequently demonstrate steep dose-response curves, suggesting the importance of full drug dosage.
Although retrospective data suggested that dose intensity may be important, several prospective randomized trials in epithelial ovarian cancer have failed to demonstrate an improved outcome by either increasing the dose of cisplatin or carboplatin per cycle or extending the duration of treatment beyond 5 to 6 cycles (6,7,8). In addition, two recent randomized studies of high-dose chemotherapy (with bone marrow or peripheral progenitor stem cell support) for advanced ovarian cancer failed to demonstrate superior survival compared to standard dose regimens (9,10).
Although evidence does not suggest dose-intensive approaches improve outcome, there is certainly a minimum dose below which survival will be compromised. Unfortunately, this dose is difficult to determine. In general, the goal should be to maintain dose intensity consistent with an acceptable toxicity profile in each individual patient. The severity of neutropenia can frequently be reduced through the administration of a bone marrow stimulatory agent (e.g., granulocyte colony stimulating factor). These drugs can be either administered at the time of documented severe bone marrow suppression or given prophylactically with chemotherapeutic regimens known to have a high risk of grade 3 or 4 myelosuppression.
Even though a commercially available agent has been shown to increase platelet counts and decrease the need for platelet transfusions, its role in the treatment of gynecologic cancers remains to be defined.
Pharmacologic Factors Influencing Treatment
Pharmacologically, it is useful to describe effective chemotherapy as concentration over time of the active agent or its metabolite at the primary site of antitumor action. Although it is not possible to determine exact pericellular pharmacokinetics, substantial information on important pharmacokinetic factors is available (11).
Drug effect = Drug concentration × Duration of exposure = C × T
Because direct measurements often are not possible, considerable focus is given to plasma concentration × time (C × T) analyses. Many important factors influence this pharmacokinetic result, including route of administration and drug absorption, transportation, distribution, biotransformation, inactivation, excretion, and interactions with other drugs.
Route of Administration and Absorption
Traditionally, drugs have been given orally, intravenously, or intramuscularly. Over the past decade, considerable attention has been given to the regional administration of chemotherapeutic agents, particularly in ovarian cancer (12,13,14,15). The intraperitoneal approach is based on the concept that the peritoneal clearance of the agent is slower than its plasma clearance and, as a result, an increased concentration of the drug in the peritoneal cavity is maintained while plasma concentrations are low.
Studies of a wide variety of chemotherapeutic agents have demonstrated a differential concentration of 30- to 1,000-fold, depending on the molecular weight, charge, and lipid solubility of the particular drug. Clinical trials in ovarian cancer have been performed with cisplatin, carboplatin, paclitaxel, and drug combinations (12). A number of reports have noted that approximately 30% of patients with ovarian cancer who have small volume residual disease after initial systemic platinum-based chemotherapy can achieve a surgically defined complete response following second-line treatment with intraperitoneal cisplatin.
Several randomized trials have now revealed that the intraperitoneal administration of cisplatin as primary therapy of small volume advanced ovarian cancer (largest tumor nodule within the peritoneal cavity ≤1 cm in maximal diameter) results in an improvement in both the time to subsequent disease progression and overall survival compared with intravenous delivery of this agent, but toxicity may be increased (13,14,15). It is anticipated that ongoing research with intraperitoneal therapy will define strategies that will reduce the side effects associated with regional drug delivery (16).
Antineoplastic agents usually produce their antitumor effect by interacting with intracellular target molecules. As a result, it is critically important that a particular drug or active metabolite be able to arrive at the cancer cell in sufficient concentration for lethal effect. After absorption, drugs may be bound to serum albumin or other blood components; their ability to penetrate various body compartments, vascular spaces, and extracellular sites is highly influenced by plasma protein binding, relative ionization at physiologic pH, molecular size, and lipid solubility.
Sanctuary Sites Unique circumstances may produce sanctuary sites, which are areas where the tumor is inaccessible to anticancer drugs and the drug concentration over time is insufficient for cell kill. Examples of such sanctuary sites include the cerebrospinal fluid and areas of large tumor masses with central tumor necrosis and low oxygen tension.
Cell Penetration Although some drugs enter the target cell by simple diffusion, in some instances cellular penetration is an active process. As an example, many of the alkylating agents depend on a carrier transport system for cellular penetration. For large macromolecules, it may be necessary for pinocytosis to accomplish cellular entry.
Many antineoplastic agents are active as intact molecules, but some require metabolism to an active form. Many of the antimetabolites require phosphorylation for cell entry. The alkylating agent cyclophosphamide requires absorption and liver metabolism to be activated. Attention to these unique metabolic requirements is needed for appropriate drug selection. For example, if direct installation of an alkylating agent is required, an agent that is active as an intact drug should be selected (e.g., Thiotepa or nitrogen mustard), rather than cyclophosphamide, because the latter drug requires hepatic biotransformation and would not be active locally. Not only is initial activation important but also the rate of metabolic degradation of the active drug or metabolite is important in determining antitumor activity. As an example, a major mechanism of drug resistance in ovarian cancer is increased metabolism of alkylating agents because of increased intracellular enzymes (e.g., glutathione-S-transferase).
Most chemotherapeutic agents are excreted through the kidney or liver. Because overall kidney or liver function is critical to normal drug excretion, it is necessary to modify the dosage of certain agents when either of these organs is functionally impaired.
Certain drugs (e.g., vincristine, doxorubicin, paclitaxel) are excreted primarily through the liver, and others (e.g., methotrexate) are excreted almost entirely by the kidney. Most experimental protocols and cooperative group trials contain formulas for dose modification for specific organ impairments that influence drug excretion.
There are multiple opportunities for clinically important drug interactions to occur during cancer treatment. These interactions may increase or decrease the antitumor activity of a particular agent, or they may increase or modify its toxicity. Types of drug interaction of potential importance include those listed in Table 3.4.
Important drug interactions with antineoplastic drugs include the following.
- The alkylating agents are highly reactive compounds and may produce direct chemical or physical inactivation when multiple drugs are mixed.
- Intestinal absorption of certain chemotherapeutic agents is altered by antibiotics that suppress bowel flora (e.g., reduced absorption of oral methotrexate), resulting in its decreased circulating level.
- Drugs such as cisplatin or methotrexate bind to albumin or plasma proteins and may be displaced from that binding by drugs that bind to similar sites, such as aspirin or sulfa, thereby increasing the circulating level of bioavailable cisplatin or methotrexate.
- Alterations in drug activation may occur, as when methotrexate increases 5-fluorouracil activation; conversely, drug interaction may antagonize antitumor effect, as when 5-fluorouracil impairs the antifolate action of methotrexate.
- Nephrotoxic antibiotics frequently alter methotrexate excretion and may increase the renal toxicity of cisplatin.
Table 3.4 Drug Interactions in Cancer Chemotherapy
Principles of Combination Chemotherapy
Combination chemotherapy has become the standard approach to the management of many adult solid tumors, including breast cancer and female pelvic malignancies. The enthusiasm for combinations results from several significant limitations inherent in single-agent chemotherapy. In addition, there is a solid theoretic basis for combination chemotherapy from a knowledge of cellular kinetics, drug metabolism, drug resistance, and tumor heterogeneity.
Limitations of Single-Drug Therapy
The major limitations of single-agent chemotherapy are the following.
- Toxicity limits the dose and duration of drug administration and thus restricts the tumor cell kill achievable.
- Adaptive mechanisms allow cell survival and eventual regrowth of resistant tumor cells in spite of lethal effects produced in the bulk of the tumor.
- Drug resistance may spontaneously develop.
- Multidrug or pleiotropic drug resistance may develop.
Several different mechanisms of resistance are seen with antineoplastic agents, and some of these are listed in Table 3.5. Most problems inherent in single-drug therapy cannot be corrected by simply altering the dose or schedule of that single drug. As a result, increasing use has been made of multidrug combination chemotherapy.
Combination Chemotherapy Mechanisms
Different chemotherapeutic agents may act in different phases of the tumor cell cycle. Use of multiple drugs with different cellular kinetic characteristics reduces the tumor mass more completely than any individual chemotherapeutic agent while minimizing the impact of singledrug resistance. For instance, if a cell cycle-nonspecific agent is administered, producing a 2-log cell kill in a tumor mass with 109 cells, and no further therapy is given, then a minor tumor response will occur, followed by tumor regrowth and no impact on survival. If a cell cycle-specific agent produces a similar degree of cell kill, only the cells coming into cell cycle will be affected by such an agent. Simply by using combinations or sequences of cell cycle-specific and -nonspecific agents, log kill can be enhanced in tumors. With identification of appropriate combinations and proper sequencing, sufficient log kill may be achieved to produce a cure.
Table 3.5 Mechanisms of Resistance to Anticancer Drugs
Combination chemotherapy can help to circumvent spontaneous mutations to drug resistance. After initial cell kill, the residual tumor may contain drug-resistant cells. The probability of the emergence of drug-resistant cells in any given population is reduced if two or more agents with different mechanisms of action can be used in a tightly sequenced treatment scheme.
Drug interactions may be additive, synergistic, or antagonistic. Combinations that result in improved therapy because of increased antitumor activity or decreased toxicity are said to be synergistic. Additive therapies produce enhanced antitumor activity equivalent to the sum of both agents acting singly. Finally, antitumor agents may actually antagonize the effect of each other, producing a lesser therapeutic effect than when used singly. For example, 5-fluorouracil prevents the antifolate action of methotrexate when used beforemethotrexate administration.
In some instances, the same drugs used in different sequences may produce a widely varied effect, suggesting the importance of schedule dependency. An example is the reduced cardiac toxicity demonstrated for weekly low-dose doxorubicin compared with high-dose bolus doxorubicin. Although schedule dependency has been an important, well-documented phenomenon in experimental tumors, its importance is less well defined for human cancer chemotherapy.
The general principles that allowed the development of successful combinations are shown in Table 3.6. Although these cannot be used in every regimen and some overlap in toxicities is common, these concepts are a central feature of most of the regimens now being used successfully in cancer treatment.
Once a treatment regimen has been selected, it is necessary to have some standardized way to evaluate the response to drug treatment. The terms complete remission (or response) and partial remission (or response) are used frequently and provide a convenient way to describe responses and compare various published regimens.
Complete Remission (Response) Complete remission (response) is the complete disappearance of all objective evidence of tumor as well as the resolution of all signs and symptoms referable to the tumor. Complete regressions of cancer are those associated in general with significant prolongation of survival.
Partial Remission (Response) A partial remission (response) has been variously defined as a 30% to 50% reduction in the size of all measurable lesions along with some degree of subjective improvement and the absence of any new lesions during therapy. Partial remissions translate in general into improved well-being for the patient but only occasionally are associated with longer overall survival.
Table 3.6 Important Factors in the Design of Drug Combinations
Finally, various terms indicate lesser responses, such as objective response or minor response, but such responses rarely result in any significant improvement in survival.
The Response Evaluation Criteria in Solid Tumors (RECIST) have been used to measure the effect of chemotherapy in an individual patient and are now used in all clinical trials(Table 3.7).
Baseline documentation of “target” and “nontarget” lesions before treatment in clinical trials is essential. All measurable lesions up to a maximum of 5 lesions per organ and 10 lesions in total, representative of all involved organs, should be identified as target lesions and recorded and measured at baseline. Target lesions should be selected on the basis of their size (lesions with the longest diameter) and their suitability for accurate repeated measurements (either clinically or by imaging techniques).
A sum of the longest diameter (LD) for all target lesions should be calculated and reported as the baseline sum LD. The baseline sum LD should be used as the reference by which to characterize the objective tumor response. All other lesions (or sites of disease) should be identified as nontarget lesions and should also be recorded at baseline. Measurements of these lesions are not required, but the presence or absence of each should be noted throughout follow-up.
Patients vary in their tolerance to chemotherapy, and tailoring the treatment to a particular patient is often necessary, particularly when treatment is administered with palliative intent. One convenient method involves the use of a “sliding scale.” A typical scheme for adjusting chemotherapy based on myelosuppression is presented in Table 3.8. Doses of myelosuppressive agents are reduced if the patient proves very sensitive to the regimen but can be returned to full levels if tolerance improves in subsequent courses.
Table 3.7 RECIST Definitions of Response
Table 3.8 Drug Dose Adjustments for Combination Chemotherapy (Sliding Scale Based on Bone Marrow Toxicity)
Many experimental protocols provide for an escalation of drug dose if no significant toxicity is experienced with initial courses of therapy. A sliding scale offers the best opportunity to give the maximum amount of therapy possible. The sliding scale presented is based only on bone marrow toxicity. If the drugs used in any particular combination have other serious toxicities, such as renal or hepatic toxicity, then sliding scales based on the other toxicities are used to minimize toxicity but maximize therapeutic effect.
Because carboplatin is cleared renally and severe marrow toxicity occasionally occurs, doseadjustment scales based on renal function have been developed (Table 3.9). Dose adjustments are based on the glomerular filtration rate (GFR) or creatinine clearance and the target serum concentration multiplied by the area under curve (AUC) or platelet nadir for the drug's antitumor activity (17). The formula is:
Dose (mg) = Target AUC × (GFR + 25)
The desired target AUC is 4 to 5 mg/mL for previously treated patients and 5 to 7 mg/mL for those previously untreated. The use of these dose-adjustment schemes tailored to the particular toxicity allows for safer administration of chemotherapeutic agents.
Antineoplastic drugs are among the most toxic agents used in modern medicine. Many of the toxic side effects, particularly those to organ systems with a rapidly proliferating cell population, are dose related and predictable. Usually, the mechanism of toxicity is similar to the mechanism that produces the desired cytocidal effect on tumors. Even organs with limited cell proliferation can be damaged by chemotherapeutic agents in either a dose-related or an idiosyncratic fashion. In almost all instances, chemotherapeutic agents are used in doses that produce some degree of toxicity to normal tissues.
Severe systemic debility, advanced age, poor nutritional status, or direct organ involvement by primary or metastatic tumor can result in unexpectedly severe side effects of chemotherapy. Idiosyncratic drug reactions also can have severe and unexpected consequences. As a result, careful monitoring of patients receiving cancer chemotherapy is a major responsibility of the treating physician.
The proliferating cells of the erythroid, myeloid, and megakaryocytic series of the bone marrow are highly susceptible to damage by many of the commonly used antineoplastic agents. Granulocytopenia and thrombocytopenia are predictable side effects of most of the commonly used antitumor agents and are seen with all effective regimens of combination chemotherapy. The severity and duration of these side effects are variable and depend on the drugs, the dose, the schedule, and the patient's previous exposure to radiation or chemotherapy.
Table 3.9 Carboplatin Dosing
In general, acute granulocytopenia occurs 6 to 12 days after administration of most myelosuppressive chemotherapeutic agents, and recovery occurs in 21 to 24 days; platelet suppression occurs 4 to 5 days later, with recovery after white cell count recovery. Several agents are unique in producing delayed bone marrow suppression, among themmitomycin C and the nitrosoureas. Marrow suppression from these drugs commonly occurs at 28 to 42 days, with recovery 40 to 60 days after treatment.
Granulocytopenia Patients with an absolute granulocyte count of less than 500/mm3 for 5 days or longer are at high risk of rapidly fatal sepsis. The wide use of empiric, broadspectrum antibiotics in febrile granulocytopenic patients with cancer has significantly decreased the likelihood of life-threatening toxicity. The importance of quickly initiating broadspectrum antibiotics in the presence of fever in a neutropenic patient, even in the absence of localizing signs of infection, cannot be overemphasized. Granulocytopenic patients should have their temperature checked every 4 hours and must be examined frequently for evidence of infection. The availability of hematopoietic growth factors has enabled physicians to reduce the duration of granulocytopenia in certain patients.
Before the initiation of antibiotics in a febrile granulocytopenic patient, cultures of possible sites of infection (e.g., blood, urine, sputum, recent surgical wound, indwelling intravenous delivery device) should be obtained. In addition, a detailed physical examination (including the throat, perianal region, and skin) should be performed, looking for a specific site of infection, which may influence the choice of antibiotic therapy (e.g., catheter infection).
Thrombocytopenia Patients with sustained thrombocytopenia who have platelet counts of less than 20,000/mm3 are at risk of spontaneous hemorrhage, particularly gastrointestinal or acute intracranial hemorrhage. Routine platelet transfusions for platelet counts below 10,000 to 20,000/mm3 have significantly reduced the risk of spontaneous hemorrhage. It is common to transfuse 6 to 10 units of donor platelets to the patient with a platelet count of less than 20,000/mm3. Repeat transfusions at intervals of 2 to 3 days for the duration of the severe thrombocytopenia are indicated. Although patients with platelet counts exceeding 50,000/mm3 do not commonly experience severe bleeding, transfusion at this level is indicated:
- if the patient manifests active bleeding
- if the patient has active peptic ulcer disease
- before and during surgical procedures
A posttransfusion platelet count performed one hour after platelet administration should show an appropriate incremental increase. If no posttransfusion platelet increase occurs, it is likely that there has been previous sensitization to random donor platelets, and the patient requires single-donor human leukocyte antigen-matched platelets for future transfusions.
Recombinant interleukin-11 can be considered for use in patients with, or anticipated to develop, severe thrombocytopenia. The drug is administered subcutaneously beginning 6 to 24 hours after chemotherapy (50 µg/kg once daily) and continued until the platelet count exceeds 50,000/mm3. Treatment with this agent should be discontinued at least 2 days before the next chemotherapy.
The gastrointestinal tract is a frequent site of serious antineoplastic drug treatment toxicity. Mucositis caused by a direct effect on the rapidly dividing epithelial mucosal cells is common; concomitant granulocytopenia allows the injured mucosa to become infected and serve as a portal of entry for bacteria and fungi into the bloodstream. Impaired cellular immunity because of underlying disease or corticosteroid therapy also can contribute to extensive infection of the gastrointestinal tract. Other side effects related to the gastrointestinal tract include impaired intestinal motility, resulting from the autonomic neuropathic effect of vinca alkaloids (vincristine and vinblastine), and nausea and vomiting,induced by many anticancer drugs.
Upper Gastrointestinal The onset of mucositis is frequently 3 to 5 days earlier than that of myelosuppression. Lesions of the mouth and pharynx are difficult to distinguish from candidiasis and herpes simplex infection. Esophagitis resulting from direct drug toxicity can be confused with radiation esophagitis or infections with bacteria, fungi, or herpes simplex because they all produce dysphagia and retrosternal burning pain. Mild oral candidiasis (thrush) responds to several oral agents. More intensive therapy will be required for esophageal or severe oral candidiasis or herpes simplex infections. Symptomatic management of painful upper gastrointestinal inflammation includes warm saline mouth rinses and topical anesthetics such as viscous lidocaine. Intravenous fluids or hyperalimentation may be required.
Lower Gastrointestinal Mucositis in the lower gastrointestinal tract is invariably associated with diarrhea. Serious complications include bowel perforation, hemorrhage, and necrotizing enterocolitis.
Necrotizing enterocolitis includes a spectrum of severe diarrheal illnesses that can be fatal in a granulocytopenic patient. Broad-spectrum antibiotic therapy may predispose the patient to necrotizing enterocolitis, as does cytotoxic chemotherapy, which can interfere with the integrity of the bowel wall. This condition is more common in patients receiving intensive chemotherapy (e.g., patients with leukemia), but it can be observed with treatment of solid tumors (e.g., gynecological malignancies). The most common organism associated with this extremely serious condition is Pseudomonas aeruginosa. Symptoms of necrotizing enterocolitis include watery or bloody diarrhea, abdominal pain, sore throat, nausea, vomiting, and fever. Physical examination usually reveals abdominal tenderness and distention. The performance of an abdominal or pelvic computed tomographic scan or ultrasound will be helpful in the evaluation of this constellation of signs and symptoms. Treatment includes the administration of broad-spectrum antibiotics with specific activity against aerobic gram-negative rods and anaerobes. Nasogastric decompression, intravenous fluids, and bowel rest may also be required. In the neutropenic patient, recovery of normal blood counts is essential for improvement of the condition. Surgical intervention is occasionally necessary.
Most anticancer drugs are capable of producing suppression of cellular and, to a lesser extent, humoral immunity. The magnitude and duration of the immunosuppression vary with the dose and schedule of drug administration and have been inadequately characterized for most chemotherapeutic agents. However, most of the acute immunosuppressive side effects do not persist after completion of drug treatment. Laboratory studies suggest a decrease in host defenses during treatment associated with a rebound to complete or nearly complete restoration 2 to 3 days after treatment is completed. This short-term immunosuppressive effect has led to increased use of intermittent chemotherapy regimens to allow immunologic recovery during courses of treatment.
Several important drug toxicities involve skin reactions. Skin necrosis and sloughing may result from extravasation of certain irritating chemotherapeutic agents such as doxorubicin, actinomycin D, mitomycin C, vinblastine, vincristine, and nitrogen mustard. The extent of necrosis depends on the quantity of drug extravasated and can vary from local erythema to chronic ulcerative necrosis. Management often includes immediate removal of the intravenous line, local infiltration of corticosteroids, ice pack therapy four times a day for 3 days, and elevation of the affected limb. Long-term monitoring of the affected area is required, and surgical debridement and full-thickness skin grafting are often necessary for severe lesions.
Alopecia is a very most common side effect. Although not intrinsically injurious, it has major emotional consequences for patients. Agents commonly associated with severe hair loss include the anthracycline antibiotics, taxanes such as paclitaxel or docetaxel, and alkylating agents such as cyclophosphamide. Most commonly used drug combinations, however, produce variable degrees of alopecia. Alopecia is reversible, and regrowth usually begins several weeks after treatment is completed. Attempts to minimize alopecia by using a variety of methods such a scalp cooling have been tried with varying degrees of success, depending on the drugs used.
Generalized allergic skin reactions can occur with chemotherapeutic agents, as they do with other drugs, and can sometimes be severe. Other skin reactions occasionally seen with chemotherapeutic agents include increased skin pigmentation (bleomycin), photosensitivity reactions, transverse banding or nail loss, folliculitis (actinomycin D, methotrexate), and radiation recall reactions (doxorubicin).
Liposomal doxorubicin, an agent demonstrated to be active in platinum-refractory ovarian cancer, can produce a painful dermatologic syndrome characterized by desquamation of the skin, most often involving the hands and feet (18). Blistering, focal or disseminated, can also be observed.
Modest elevations in aminotransferase, alkaline phosphatase, and bilirubin levels are frequently seen with many anticancer agents, but they resolve soon after treatment is completed. Nevertheless, more severe reactions do occur. Long-term administration of methotrexate induces hepatic fibrosis that can progress to cirrhosis. The cirrhosis and drug-induced hepatitis should be managed by withdrawal of the toxic agent, with the same supportive measures that are used for hepatitis or cirrhosis of any cause.
Preexisting liver disease or exposure to other hepatotoxins may increase the risk. Antimetabolites, such as 6-mercaptopurine and 6-thioguanine, can produce reversible cholestatic jaundice. Transient liver enzyme abnormalities are seen with cytosine arabinoside, the nitrosoureas, and L-asparaginase. Mithramycin, an agent occasionally used to control hypercalcemia, frequently causes marked elevations in liver enzyme levels associated with clotting disorders and renal insufficiency.
Patients with cancer have a wide variety of problems that can manifest as pulmonary complications. Respiratory compromise resulting from lung metastases, pulmonary emboli, radiation pneumonitis, tumor-induced neuromuscular dysfunction, and pneumonia all may be significant complications. In addition, direct pulmonary toxicity from commonly used anticancer drugs is sometimes seen.
Interstitial Pneumonitis Interstitial pneumonitis with pulmonary fibrosis is the usual pattern of lung damage associated with cytotoxic drugs. Agents likely to cause such an effect arebleomycin, alkylating agents, gemcitabine, and the nitrosoureas. The physical and chest radiologic findings are not easily distinguishable from those of interstitial pneumonitis resulting from infectious agents, viruses, or lymphangitic spread of cancer.
Management of drug-induced interstitial pneumonitis includes discontinuation of the suspected agent and supportive care. Steroids may be of symptomatic benefit in some patients.
Cardiac toxicity is seen with several important cancer chemotherapeutic agents. Although the myocardium consists of largely nondividing cells, drugs of the anthracycline antibiotic class—specifically, doxorubicin and daunomycin—can cause severe cardiomyopathy.
The risk of cardiac toxicity increases with the total cumulative dose of doxorubicin. For this reason, a cumulative dose of 500 mg/m2 of ideal body surface area is now widely used as the maximum tolerable dose of doxorubicin. With careful and frequent monitoring of left ventricular function by means of ejection fraction studies, therapy can be continued to higher doses if no satisfactory alternative exists. More infrequently, anthracyclines and paclitaxel can cause acute arrhythmias that usually disappear within a few days of drug treatment. They appear not to be related to total drug dose. Anthracycline cardiac toxicity is potentiated by radiation.
The medical management of cardiomyopathy induced by anthracyclines is supportive but usually unsatisfactory. Early detection of cardiac compromise with radionuclide cardiac scintigraphy before the clinical manifestations of congestive heart failure appear is important. Discontinuation of the drug at the first indication of decreasing left ventricular function minimizes the risk of cardiovascular decompensation.
Rarely, cyclophosphamide has been reported to produce cardiotoxicity, particularly in the massive doses used in conjunction with bone marrow transplantation. With conventional doses of cyclophosphamide, this complication is unlikely. Busulfan and mitomycin C have been reported to cause endocardial fibrosis and myocardial fibrosis, respectively. In some patients, 5-fluorouracil has been reported to be a rare cause of angina pectoris.
Cardiac toxicity has also been recognized as an important side effect of trastuzumab, a novel targeted therapy (HER-2/neu receptor) commonly used in the management of breast cancer, especially when the agent is delivered with doxorubicin or paclitaxel (19).
A major toxicity of the antiangiogenic agent bevacizumab is the development of hypertension (20). In a patient with preexisting cardiac abnormalities, this effect has the potential to result in deterioration of heart function. This concern is likely to increase as this class of agents is increasingly used in routine clinical practice in the management of gynecologic malignancies.
In addition to chemotherapeutic agents, various other cancer-related complications may produce chronic azotemia or acute renal failure, including fluid depletion, infection, tumor infiltration of the kidney, ureteral obstruction by tumor, radiation damage, and tumor lysis syndrome.
Drugs that cause kidney damage include:
- cisplatin, which produces renal tubular toxicity associated with azotemia and magnesium wasting
- methotrexate, which can precipitate in the renal tubules, causing oliguric renal failure (methotrexate toxicity can be prevented by maintenance of a high urine volume and alkalinization of the urine)
- nitrosoureas, which cause a chronic interstitial nephritis with chronic renal failure
- mitomycin C, which causes a systemic microangiopathic hemolysis and acute renal failure
Metabolites of cyclophosphamide are irritants to the bladder mucosa and cause a chronic hemorrhagic cystitis, particularly during high-dose or prolonged treatment. Vigorous hydration and diuresis can reduce the risk of this complication.
Treatment of drug-related genitourinary toxicity requires discontinuation of the possibly nephrotoxic drugs and volume expansion to increase glomerular filtration. Specific metabolic abnormalities, such as hyperuricemia and hypomagnesemia, should be corrected. If oliguria develops or if medical management is unsuccessful in restoring acceptable kidney function, then short-term peritoneal dialysis or hemodialysis may be required. Daily administration of 3 L of fluid containing 100 to 150 mEq of sodium bicarbonate per liter maintains the urinary pH above 7. Because methotrexate is poorly dialyzed, prolonged toxic levels can result if leucovorin rescue therapy is not continued until the methotrexateconcentration is less than 5 × 10-8 M.
N-acetylcysteine or mesna (sodium mercaptoethanesulfonate) has been used in conjunction with very high doses of cyclophosphamide or ifosfamide to prevent bladder toxicity by inactivating the toxic metabolite (acrolein). Persistent hemorrhagic cystitis that does not respond to conservative management may be treated with ε-aminocaproic acid.
Many antineoplastic drugs are associated with some central or peripheral neurotoxicity. These neurologic side effects usually are mild, but occasionally they can be severe.
Vinca Alkaloids The vinca alkaloids (vincristine, vinblastine, and vindesine) are commonly associated with peripheral motor, sensory, and autonomic neuropathies, which are the major side effects of vincristine. Toxicity first appears as loss of deep tendon reflexes with distal paresthesias. Cranial nerves can be affected, and the autonomic neuropathy can appear as adynamic ileus, urinary bladder atony with retention, or hypotension. All of these neurologic toxicities from the vinca alkaloids are slowly reversible after cessation of the offending drug.
Cisplatin Cisplatin produces ototoxicity, peripheral neuropathy, and, rarely, retrobulbar neuritis and blindness. High doses of cisplatin, which may be used in ovarian cancer therapy, are particularly likely to produce a progressive and somewhat delayed peripheral neuropathy. This defect is characterized by sensory impairment and loss of proprioception, whereas motor strength usually is preserved. Progression of this neuropathy 1 to 2 months after cessation of high-dose cisplatin has been reported.
Paclitaxel Paclitaxel is associated with the development of a peripheral sensory neuropathy. The incidence and severity of symptoms relate to the peak levels of the agent reached in the plasma. In addition, the combination of paclitaxel and cisplatin (or carboplatin) has the potential to be more neurotoxic than either agent used alone (21).
Other Drugs Rarely, 5-fluorouracil can be associated with an acute cerebellar toxicity, apparently related to its metabolism to fluorocitrate, a neurotoxic metabolite of the parent compound. Hexamethylmelamine has been reported to produce peripheral neuropathy and encephalopathy. Some improvement in the peripheral neuropathy has been reported with administration of B vitamin supplements, but therapeutic effectiveness may be reduced.
Vascular and Hypersensitivity Reactions
Occasionally, severe hypersensitivity reactions in the form of anaphylaxis develop with chemotherapeutic agents. In rare cases, this has been associated with cyclophosphamide, doxorubicin, cisplatin, intravenous melphalan, and high-dose methotrexate. Bleomycin administration may be associated with marked fever reactions, anaphylaxis, Raynaud's phenomenon, and a chronic scleroderma-like reaction. The same reactions have been reported with procarbazine, etoposide, and teniposide.
Hypersensitivity reactions have been seen with paclitaxel and are believed to result from hypersensitivity to the cremophor vehicle. They can be ameliorated with intravenous infusions of dexamethasone (20 mg), diphenhydramine (50 mg), and cimetidine (300 mg) 30 minutes before paclitaxel is administered. A similar incidence of hypersensitivity reactions is observed with docetaxel, a closely related antineoplastic agent to paclitaxel.
Carboplatin and cisplatin may be associated with a significant risk of hypersensitivity reactions in patients who have been treated with more than six total courses of a platinum agent (22).
Many antineoplastic agents are mutagenic and teratogenic. The potential of these agents to induce second malignancies appears to vary with the class of agent (23). Alkylating agents (especially melphalan), procarbazine, and the nitrosoureas seem to be the major offenders. Prolonged use of etoposide has also been associated with the development of leukemia.
The cumulative 7-year risk of acute nonlymphocytic leukemia developing in patients treated primarily with oral melphalan for ovarian cancer is as high as 9.6% in patients receiving therapy for more than 1 year (24). Although cisplatin has also been suggested to be associated with the development of acute leukemia, the risk is lower than with the alkylating agents (25).
Evidence from long-term studies of Hodgkin's disease suggests a major risk with combined chemotherapy and radiation therapy. In such patients, there is a risk of acute leukemia as well as an increase in solid tumors, seen particularly in the radiation ports. An increase in the frequency of acute leukemia has been reported in patients treated for Hodgkin's disease, multiple myeloma, and ovarian cancer. The second malignancy commonly occurs 4 to 7 years after successful therapy. Encouragingly, evidence suggests that after 11 years, the risk of acute leukemia in patients treated for Hodgkin's disease decreases to that of the normal population.
The long-term follow-up of women cured of choriocarcinoma, primarily with antimetabolite therapy, reveals no evidence of an increased risk of second malignancy. Radiation alone also appears to produce a relatively low risk of late leukemia, as do chemotherapeutic regimens alone, particularly those without alkylating agents or procarbazine. Combination chemotherapy (including cisplatin-based treatment of ovarian cancer) and limited-field radiation therapy increase the risk only slightly.
Particularly high risks are associated with:
- extensive radiation therapy plus combination chemotherapy
- prolonged alkylating agent therapy (longer than 1 year)
- age older than 40 years at initial treatment
Many cancer chemotherapeutic agents have profound and lasting effects on testicular and ovarian function. Chemotherapeutic agents, particularly alkylating agents, can cause azoospermia and amenorrhea. Secondary sexual characteristics related to hormonal function usually are less disturbed. Prolonged intensive combination chemotherapy commonly produces azoospermia in men, and recovery is uncommon.
The onset of amenorrhea and ovarian failure is accompanied by an elevation of the serum follicle-stimulating hormone and luteinizing hormone and a decrease in the serum estradiol level. Occasionally, this hormonal pattern can be seen before the onset of amenorrhea. If the characteristic pattern is seen, patients should be advised to consider conception because these findings predict premature ovarian failure and early menopause.
When short-term intensive chemotherapy is used, particularly with antimetabolites, vinca alkaloids, or antitumor antibiotics, injury to the reproductive system is less common. For example, men treated for testicular cancer, children with acute leukemia, and women cured of gestational trophoblastic disease or ovarian germ cell malignancies usually have recovered reproductive capacity after therapy (26,27,28,29).
Chemotherapy in Pregnancy
Risk of congenital abnormalities from chemotherapeutic agents is highest during the first trimester, especially when antimetabolites (e.g., cytosine arabinoside or methotrexate) and alkylating agents are used. Chemotherapy administered during the second or third trimesters usually is not associated with an increase in fetal abnormalities, although the number of patients studied is relatively small (see Chapter 17).
Inappropriate Antidiuretic Hormone Secretion Inappropriate antidiuretic hormone secretion is characterized by hyponatremia, high urine osmolality, and high urinary sodium values, and it is associated with several malignancies, most commonly small cell carcinoma of the lung. It can also be seen as a complication of vinca alkaloids. Symptoms are primarily neurologic and include altered mental status, confusion, lethargy, seizures, and coma. The severity of symptoms is related to the rapidity of development of hyponatremia. The diagnosis rests on:
- the documentation of hyponatremia
- the presence of a urine that is hypertonic to plasma
- the exclusion of hypothyroidism or adrenal insufficiency
Hyperuricemia Hyperuricemia may be a complication of effective cancer chemotherapy in certain tumors, particularly hematologic malignancies in which rapid tumor lysis is seen in response to initial treatment. Rapid tumor lysis releases predominant intracellular ions and uric acid and can result in life-threatening hyperkalemia, hyperphosphatemia, hypocalcemia, and hyperuricemia. Renal failure associated with hyperuricemia can be severe. Prevention of the tumor lysis syndrome requires maintenance of a high urinary output, maintenance of high urinary pH (above 7.0), and prophylactic use of the xanthine oxidase inhibitor allopurinol.
This class of antineoplastic agent acts primarily by chemically interacting with DNA. These drugs form extremely unstable alkyl groups that react with nucleophilic (electron-rich) sites on many important organic compounds such as nucleic acids, proteins, and amino acids. These interactions produce the primary cytotoxic effects.
Alkylating agents commonly bind to the N-7 position of guanine and to other key DNA sites. In doing so, they interfere with accurate base pairing, crosslink DNA, and produce single- and double-stranded breaks. This results in the inhibition of DNA, RNA, and protein synthesis.
Table 3.10 Alkylating Agents Used for Gynecologic Cancer
Because some effects of alkylating agents are similar to those of irradiation, these drugs are often called radiomimetic. Most of the effective alkylating agents are bifunctional or polyfunctional, and have two or more potentially unstable alkyl groups per molecule. These bifunctional alkylating agents allow crosslinkage of DNA that results in cellular disruption.
Because all alkylating agents have similar mechanisms of action, there tends to be crossresistance to other agents of the same class.
Although several hundred alkylating agents exist, those most commonly in use include cyclophosphamide, melphalan, Thiotepa, chlorambucil, and ifosfamide.
In addition to the more common alkylating agents, several antineoplastic agents of different types are usually classified as alkylating-like agents, although their precise mechanism of action is less well understood and is probably not exclusively alkylation. These include the nitrosoureas, DTIC (dacarbazine), and the platinum analogs cisplatin and carboplatin.
The characteristics of the commonly used alkylating agents are listed in Table 3.10, and the alkylating-like agents are listed in Table 3.11.
The antitumor antibiotics are antineoplastic drugs that, in general, have been isolated as natural products from fungi found in the soil. These natural products usually have extremely complex and different chemical structures, although they function in general by forming complexes with DNA.
The interaction between these drugs and DNA often involves intercalation: The compound is inserted between DNA base pairs. A second mechanism thought to be important in their antitumor action is the formation of free radicals capable of damaging DNA, RNA, and vital proteins. Other effects include metal ion chelation and alteration of tumor cell membranes. This class of antineoplastic agents is thought to be cell cycle-nonspecific.
Major drugs in this family include the anthracycline antibiotics doxorubicin, liposomal doxorubicin, and daunorubicin, as well as actinomycin D, bleomycin, mitomycin C, andmithramycin.
Anthracyclines The anthracyclines are antibiotics isolated from the fungi Streptomyces.
These pigmented compounds have an anthraquinone nucleus attached to an amino sugar and have multiple mechanisms of action. Because of the planar structure of the anthraquinone moiety, these agents act as intercalators in the DNA double helix. In addition, they are known to chelate divalent cations and are avid calcium binders. These agents cause single-stranded DNA breaks, inhibit DNA repair, and actively generate free radicals that are capable of producing DNA damage. Anthracyclines are capable of reacting directly with cell membranes, disrupting membrane structure, and altering membrane function.
Table 3.11 Alkylating-Like Agents Used for Gynecologic Cancer
Bleomycin Bleomycin was also isolated from the Streptomyces fungus. Its structure contains a DNA-binding fragment and an ion-binding unit. It appears to produce its antitumor action primarily by producing single- and double-stranded breaks in DNA, mainly at sites of guanine bases. The drug is primarily excreted in the urine, and increased toxicity may be seen in patients with impaired renal function.
Mitomycin C Mitomycin C is another antibiotic that was isolated from the Streptomyces fungus. It is activated in vivo into an alkylating agent that can bind DNA, producing crosslinks and inhibiting DNA synthesis. In addition, it has a quinone moiety that can generate free radical reactions similar to those seen with the anthracycline antibiotics. It is administered intravenously and is degraded primarily by metabolism. Renal clearance is not a major mechanism of excretion.
Mithramycin Mithramycin is an antitumor antibiotic isolated from another Streptomyces species. It has intrinsic antitumor properties and is also effective in the management ofhypercalcemia. Its primary mechanism of action seems to be the inhibition of RNA synthesis, although it binds to DNA and inhibits DNA and protein synthesis.
Some of the important characteristics of the antitumor antibiotics are listed in Table 3.12.
Table 3.12 Antitumor Antibiotics Used for Gynecologic Cancer
The antimetabolite family of antineoplastic agents interacts with vital intracellular enzymes, leading to their inactivation or to the production of fraudulent products incapable of normal intracellular function. In general, their structures resemble analogs of normal purines and pyrimidines, or they resemble normal substances that are vital for cell function. Some antimetabolites are active as intact drugs, and others require biotransformation to active agents.
Although many of these agents act at different sites in biosynthetic pathways, they appear to exert their antitumor activity by disrupting functions crucial to the viability of the cell. These effects are usually more disruptive to actively proliferating cells; thus, the antimetabolites are classed in general as cell cycle-specific agents.
Although hundreds of antimetabolites have been investigated in cancer treatment, only a few are commonly used. They include:
- the folate antagonist methotrexate, which inhibits the enzyme dihydrofolate reductase
- the purine antagonists 6-mercaptopurine and 6-thioguanine
- the pyrimidine antagonists 5-fluorouracil (5-FU) and cytosine arabinoside
- the ribonucleotide reductase inhibitor hydroxyurea
- the nucleoside analog gemcitabine
In most instances, the antimetabolites are used not as single drugs but in combinations because of their cell cycle specificity and their capacity for complementary inhibition. Antimetabolites commonly used in the treatment of gynecologic malignancies are summarized in Table 3.13.
The most common plant alkaloids in use are the vinca alkaloids, natural products derived from the common periwinkle plant (Vinca rosea), although the epipodophyllotoxins andpaclitaxel are used frequently in gynecologic malignancies (Table 3.14). Like most natural products, these compounds are large and complex molecules, but vincristine andvinblastine differ only by a single methyl group on one side chain.
Vincristine and vinblastine act primarily by binding to vital intracellular microtubular proteins, particularly tubulin. Tubulin binding produces inhibition of microtubule assembly and destruction of the mitotic spindle, and cells are arrested in mitosis. In general, this class of antineoplastic agent is believed to be cell cycle-specific. At high concentrations, these drugs also have effects on nucleic acid and protein synthesis.
Table 3.13 Antimetabolites Used for Gynecologic Cancer
Table 3.14 Plant Alkaloids
Paclitaxel has a unique mechanism of action. It binds preferentially to microtubules and results in their polymerization and stabilization. Paclitaxel-treated cells contain large numbers of microtubules, free and in bundles that result in disruption of microtubular function and, ultimately, cell death. Renal clearance is minimal (5%).
Vinblastine is used primarily in the treatment of ovarian germ cell tumors. Its primary toxicity is myelosuppression. In contrast, vincristine causes little myelosuppression. Its primarydose-limiting toxicity is peripheral neuropathy. Vincristine has been used in the treatment of cervical carcinoma.
A second family of plant alkaloids has been documented to have significant antitumor properties. Members of this family, known as the epipodophyllotoxins, are extracts from the mandrake plant. Although the primary plant extracts had tubulin-binding properties similar to those of the vinca alkaloids, the active derivative, etoposide, does not seem to function either by inhibiting mitotic spindle formation or by tubulin binding. Rather, the drug appears to function by causing single-stranded DNA breaks. Unlike many of the other compounds that act primarily by DNA interactions, the agent appears to be cell cycle-specific and scheduledependent. The dose-limiting toxicity is myelosuppression. Other toxicities include an infusion rate-limited hypotension, nausea, vomiting, anorexia, and alopecia.
Paclitaxel is a complex agent in the class of drugs known as taxanes. Its major toxic effects include bone marrow suppression, alopecia, myalgias, arthralgias, and hypersensitivity reactions (29). The most common dose-limiting toxicity is granulocytopenia, although with certain schedules the limiting toxicity is peripheral sensory neuropathy. The drug is active in cancers of the ovary, endometrium, cervix, and breast.
Table 3.15 Topoisomerase-1 Inhibitors
A second taxane, docetaxel, is also active in cancers of the ovary, endometrium, and breast (30). The dose-limiting toxicity of docetaxel is bone marrow suppression, principally neutropenia. Hypersensitivity reactions are also observed.
This class of antineoplastic agent exerts its cytotoxic effect through inhibition of the enzyme topoisomerase-1 (Table 3.15) (31). This is a critically important enzyme in DNA replication, repair, and transcription. Topoisomerase-1 inhibitors bind to the enzyme-DNA complex, leading to permanent strand breaks and cell death.
Topotecan, the first topoisomerase-1 inhibitor approved for clinical use in the United States, is active in platinum-refractory ovarian cancer and cervical cancer. The major toxicity of the agent is bone marrow suppression. The drug was developed for administration on a 5-day schedule but is frequently utilized on an more convenient weekly schedule. An oral formulation of the drug is currently undergoing investigation.
Irinotecan, a second topoisomerase-1 inhibitor, has revealed activity in both cancers of the ovary and cervix (31). The major side effects of the agent are bone marrow suppression and diarrhea.
There are additional agents that are employed in the management of gynecologic malignancies (Table 3.16).
Of particular interest are a group of agents currently classified as being antiangiogenic drugs (e.g., bevacizumab) that appear to exert their biological effects on either the normal or abnormal blood vessels delivering nutrients to the malignancy (20,32). To date, experience with this class of drugs in the management of gynecologic malignancies is limited, but existing data suggest these agents may ultimately play an important role in routine disease management.
New Drug Trials
A number of chemotherapeutic agents have been studied experimentally but are not commercially available. Many of these agents have already demonstrated activity against human tumors, but sufficient evidence to allow human experimentation has not yet been acquired. In addition, many investigational agents are being studied in phase I and phase II trials.
Table 3.16 Miscellaneous Agent
Phase I Trials These studies define the spectrum of toxicity of a new chemotherapeutic agent and are complete when the dose-limiting toxicity of any particular dose and schedule has been defined.
Phase II Trials These studies usually use the dose established from phase I trials and apply this dose and schedule to selected tumor types of importance.
Phase III Trials These studies compare one effective treatment with another in a randomized fashion.
- Skipper HE, Schabel FM Jr, Mullett LB. Implications of biochemical, cytokinetic, pharmacologic, and toxicologic relationships in the design of optimal therapeutic schedules.Cancer Chemother Rep 1950;54:431-450.
- Goldie JH, Coldman AJ. A mathematical model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 1979;63:1727-1733.
- Norton L, Simon R. Predicting the course of Gompertzian growth. Nature 1976;264:542-544.
- Norton L, Simon R. The Norton-Simon hypothesis revisited. Cancer Treat Rep 1986;70:163-169.
- Ling V. Drug resistance and membrane alteration in mutants of mammalian cells. Can J Genet Cytol 1975;17:503-515.
- Hakes TB, Chalas E, Hoskins WJ, et al. Randomized prospective trial of 5 versus 10 cycles of cyclophosphamide, doxorubicin, and cisplatin in advanced ovarian carcinoma.Gynecol Oncol 1992; 45:284-289.
- McGuire WP, Hoskins WJ, Brady MS, Homesley HD, Creasman WT, Berman LM, et al. Assessment of dose-intensive therapy in suboptimally debulked ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol 1995;13:1589-1599.
- Gore M, Mainwaring P, A'Hern R, MacFarlane V, Slevin M, Harper P, et al. Randomized trial of dose-intensity with singleagent carboplatin in patients with epithelial ovarian cancer. J Clin Oncol 1998;16:2426-2434.
- Mobus V, Wandt H, Frickhofen N, et al. Phase III trial of highdose sequential chemotherapy with peripheral blood stem cell support compared with standard dose chemotherapy for first-line treatment of advanced ovarian cancer: Intergroup trial of the AGO-Ovar-AIO and EBMT. J Clin Oncol 2007;25:4187-4193.
- Grenman S, Wiklund T, Jalkanen J, et al. A randomized phase III study comparing high-dose chemotherapy to conventionally dosed chemotherapy for stage III ovarian cancer: the Finnish Ovarian Cancer (FINOVA) study. Eur J Cancer 2006;42:2196-2199.
- Chu E, DeVita VT Jr. Principles of Medical Oncology. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: principles and practice of oncology, 7th ed. Philadelphia: Lippincott-Raven Publishers, 2005:295-306.
- Markman M. Intraperitoneal antineoplastic drug delivery: rationale and results. Lancet Oncol 2003;4:277-283.
- Alberts DS, Liu PY, Hannigan EV, O'Toole R, Williams SD, Young JA, et al. Intraperitoneal cisplatin plus intravenous cyclophosphamide versus intravenous cisplatin plus intravenous cyclophosphamide for stage III ovarian cancer. N Engl J Med 1996;335:1950-1955.
- Markman M, Bundy BN, Alberts DS, Fowler JM, Clark-Pearson DL, Carson LF, et al. Phase III trial of standard-dose intravenous cisplatin plus paclitaxel versus moderately high-dosecarboplatin followed by intravenous paclitaxel and intraperitoneal cisplatin in small-volume stage III ovarian carcinoma: an intergroup study of the Gynecologic Oncology Group, Southwestern Oncology Group, and Eastern Cooperative Oncology Group. J Clin Oncol 2001;19: 1001-1007.
- Armstrong DK, Bundy B, Wenzel L, et al. Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med 2006;354:34-43.
- Markman M, Walker JL. Intraperitoneal chemotherapy of ovarian cancer: a review, with a focus on practical aspects of treatment. J Clin Oncol 2006;24:988-994.
- Calvert AH, Newell DR, Gumbrell LA, O'Reilly S, Burnell M, Boxall FE, et al. Carboplatin dosage: prospective evaluation of a simple formula based on renal function. J Clin Oncol1989;7: 1748-1756.
- Muggia FM, Hainsworth JD, Jeffers S, Miller P, Groshen S, Tan M, et al. Phase II study of liposomal doxorubicin in refractory ovarian cancer: antitumor activity and toxicity modification by liposomal encapsulation. J Clin Oncol 1997;15:987-993.
- Perez EA, Rodeheffer R. Clinical cardiac tolerability of trastuzumab. J Clin Oncol 2004;22:322-329.
- Burger RA, Sill MW, Monk BJ, Greer BE. Sorosky JI. Phase II trial of bevicizumab in persistent or recurrent epithelial ovarian cancer or primary peritoneal cancer: A Gynecologic Oncology Group study. J Clin Oncol 2007;25:515-5171.
- Connelly E, Markman M, Kennedy A, Webster K, Kulp B, Peterson G, et al. Paclitaxel delivered as a 3-hr infusion with cisplatin in patients with gynecologic cancers: unexpected incidence of neurotoxicity. Gynecol Oncol 1996;62:166-168.
- Markman M, Kennedy A, Webster K, Elson P, Peterson G, Kulp B, et al. Clinical features of hypersensitivity reactions to carboplatin. J Clin Oncol 1999;17:1141-1145.
- Van Leeuwen FE, Travis LB. Second cancers. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: principles and practice of oncology, 7th ed. Philadelphia: Lippincott-Raven Publishers, 2005:2575-2601.
- Greene MH, Boice JD Jr, Greer BE, Blessing JA, Dembo AJ. Acute nonlymphocytic leukemia after therapy with alkylating agents for ovarian cancer: a study of five randomized clinical trials. N Engl J Med 1982;307:1416-1421.
- Travis LB, Holowaty EJ, Bergfeldt K, Lynch CF, Kohler BA, Wiklund T, et al. Risk of leukemia after platinum-based chemotherapy for ovarian cancer. N Engl J Med 1999;340:351-357.
- Bower M, Newlands ES, Holden L, Short D, Brock C, Rustin GJS, et al. EMA/CO for high-risk gestational trophoblastic tumors: results from a cohort of 272 patients. J Clin Oncol1997;15: 2636-2643.
- Brewer M, Gershenson DM, Herzog CE, Mitchell MF, Silva EG, Wharton JT. Outcome and reproductive function after chemotherapy for ovarian dysgerminoma. J Clin Oncol1999;17:2670-2675.
- Tangir J, Zelterman D, Ma W, Schwartz PE. Reproductive function after conservative surgery and chemotherapy for malignant germ cell tumors of the ovary. Obstet Gynecol2003;101:251-257.
- Low JJ, Perrin LC, Crandon AJ, Hacker NF. Conservative surgery to preserve ovarian function in patients with malignant ovarian germ cell tumors: a review of 74 cases. Cancer2000;89:391-398.
- Gelmon K. The taxoids: paclitaxel and docetaxel. Lancet 1994; 344:1267-1272.
- Pizzolato JF, Saltz LB. The camptothecins. Lancet 2003;361: 2235-2242.
- Cannistra SA, Matulonis UA, Person RT, et al. Phase II study of bevicizumab in patients with platinum-resistant ovarian caner or peritoneal serous cancer. J Clin Oncol 25:5180-5186.