The addition of chemotherapy to regimens that previously utilized only surgery and/or radiation has significantly improved the survival for children with cancer. Radiation is a mainstay of adjuvant therapy for adults with brain tumors; however, the long-term sequelae of brain irradiation in young children make the use of chemotherapy to delay or reduce the need for radiation particularly appealing. However, several factors must be considered when incorporating chemotherapeutic agents into the treatment of brain tumors. This chapter discusses some general principles of the administration of chemotherapy, with emphasis placed on those aspects that are particularly important in the treatment of children with brain tumors.
Chemotherapy is a critical component in the treatment of many pediatric brain tumors. It has been used to improve cure rates, but also has been effective at reducing the intensity of or delaying radiation, which can translate into a significant reduction in long-term adverse effects. This strategy has been particularly effective at reducing and delaying the need for radiation therapy, with a subsequent correlative improvement in cognitive and other deficits in some tumors.1,2 In general, chemotherapy can be successful at reducing tumor volume, potentiating therapies such as radiotherapy, and preventing relapse by eradicating micrometastatic and residual disease. However, it is most effective when partnered with local control by either radiation or surgery.
The choice of chemotherapy and method or timing of administration can vary widely. Although each agent has activity specific to a particular tumor, innate tumor resistance has led to common usage of combination chemotherapy in an attempt to overcome drug resistance. In addition, evidence demonstrating multiple cell subpopulations in a given tumor suggests that multiple chemotherapy agents may be required to optimally address various cell groups.3,4 However, although combination chemotherapy can inhibit tumor growth more successfully than single agents, overlapping side effects can exponentially worsen the associated toxicity. Thus, optimal chemotherapy combinations rely on nonoverlapping toxicity profiles and dissimilar mechanisms of action to eradicate tumor cells.
Dosing Schemas: Timing
Chemotherapy administered prior to local surgical control is termed neoadjuvant. The goal for neoadjuvant therapy is often to shrink the tumor, which can translate into a superior re-section with decreased morbidity or a smaller radiation field.5,6 The tumor′s response, both radiographic and histological, to neoadjuvant therapy may be predictive of eventual therapeutic response, providing information that may guide further therapy and more accurately predict outcome.7 Also, in an era of increasing focus on targeted therapies, neoadjuvant timing can provide an in vivo assessment of pathway inhibition by enabling histological assessment after initial exposure to therapy. Finally, neoadjuvant therapy is thought to treat micrometastatic disease, against which surgery and radiation are ill-suited. In clinical scenarios where radiation is an effective treatment tool, neo adjuvant chemotherapy can delay radiation, allowing some mitigation of radiation-induced cognitive deficits, However, if the neoadjuvant therapy is only somewhat effective, then delaying more definitive therapy such as radiation or surgery may ultimately have a negative impact on overall survival.8
Some chemotherapy may be given at the time of surgical re-section. Although this is by definition a time-limited exposure, it can provide several advantages in the treatment of brain tumors. First, residual cells in the tumor bed can be eradicated when the tumor burden is minimal. Local instillation at this time also has favorable chemotherapy distribution to surfaces most at risk for residual tumor. Intraoperative administration can provide a dose-intensive drug exposure that may not otherwise be achievable without severe toxicities, completely bypassing the blood–brain barrier as discussed later.
Adjuvant therapy, given after surgical resection, comprises the most commonly used chemotherapy schedule. This is especially true in central nervous system (CNS) tumors, which often require immediate surgical resection for amelioration of symptoms. The goal of adjuvant therapy is to complete treatment of minimal residual disease left after maximal surgical resection by direct inhibition or killing of cycling tumor cells. Use of adjuvant therapy can thus lower the risk of recurrence or progression.
Unfortunately, many brain tumors do not respond to standard doses of chemotherapy, which has led to the use of high-dose chemotherapy in an attempt to intensify treatment. This has been effective in several tumor types including infant tumors,9–11 although it has been disappointing in other subtypes, including recurrent tumors.12,13 In addition, because chemotherapy targets any rapidly dividing cells, many normal cells and tissues are affected in addition to malignant cells. As therapy intensifies to provide greater tumoral exposure, overcome tumor cell resistance, and increase CNS penetration, adverse effects on bone marrow and other organ systems worsen as well. High-dose chemotherapy can thus lead to prolonged marrow suppression and increased risk of a life-threatening infection. One strategy to protect patients who have undergone high-dose chemotherapy is to provide an autologous stem cell rescue, which enables more rapid marrow recovery and consequently a decreased risk of life-threatening infections. Increasing the dose intensity of chemotherapy with stem cell rescue has been effective in a variety of pediatric brain tumors,14,15 although generally only in the context of minimal residual disease.16,17
In direct contrast, drugs may be given on a metronomic schedule, which is based on more frequent, lower dose administration. Because most classes of chemotherapy agents damage cells during active division, this approach theoretically can interdict a greater proportion of tumor cells due to intermittent proliferation as well as inhibit endothelial cell division (and thus angiogenesis).18,19 Other research has demonstrated modulation of the immune system as a potential mechanism for the positive effects of the metronomic schedule.20 One clear advantage of this strategy is significantly improved tolerability as well as a theoretically decreased risk of the development of resistant clones.
Radiation therapy is an effective therapy for many CNS tumors. However, its utility is limited by both inherent tumor resistance as well as negative sequelae associated with radiation. Resistance to radiation is multifactorial, caused by quiescence of some tumor cells, support from the surrounding micro-environment, and activation of radiation repair mechanisms.21 Sequelae include cognitive impairment, endocrine abnormalities, second malignancy, and vasculopathy with stroke.2 Thus, numerous attempts have been made to render tumor tissue more sensitive to radiation, enabling an improvement in tumor response or a decrease in the required total dose of radiation therapy. Chemotherapy has been studied in this context, striving to remodel the microenvironment to prevent the support of tumor growth22 and to inhibit repair mechanisms. Traditional chemotherapy agents in low-doses have been used to radiosensitize tumors without adding to radiation toxicity.23 Newer, targeted drugs have also been used in an attempt to heighten sensitivity by impairing DNA repair via mTOR inhibition,24 PARP inhibition, or by other mechanisms25,26 ( Table 14.1).
One of the most important factors governing the penetration of systemically delivered chemotherapy into the CNS and brain tumors is the blood–brain barrier (BBB). Many chemotherapeutic agents have minimal to no penetration across an intact BBB. Although delivery of chemotherapy into CNS tumors may be permitted by disruption of the BBB due to abnormal tumor vasculature or following surgery or radiation, selection of agents with sufficient CNS penetration must be considered to adequately treat areas of residual disease.27
Factors that contribute to the CNS penetration of an agent are its molecular structure and size, degree of ionization, degree of protein binding, lipophilicity and hydrophilicity, and whether the agent is a substrate for drug efflux transporters such as P-glycoprotein (PGP) or multidrug resistance protein (MRP) family transport proteins.28 Drugs used in the treatment of CNS diseases are often small, nonpolar, lipophilic drugs that are able to enter the CNS by passive diffusion, whereas larger, polar, or hydrophilic agents often require active transport mechanisms and thus in general have decreased CNS penetration.27 Although selection of agents that have good CNS penetration is important, alternative methods to optimize drug delivery have also been utilized.
There are several approaches to the administration of chemotherapy that attempt to overcome or disrupt the BBB. One such approach is the use of high-dose chemotherapy with stem cell rescue, as described above. This method overcomes the issue of unfavorable plasma to CNS concentration ratios by giving very high doses of chemotherapy, thus achieving high systemic concentrations and therefore an adequate level in the CNS. These doses would result in an excessive risk of sepsis due to prolonged myelosuppression if administered without stem cell support. Another method that has been studied is the coadministration of chemotherapy with agents to disrupt the BBB. Agents for BBB disruption include anesthetics, mannitol, bradykinin analogues, ultrasound, and radiation.29–33 Perhaps the most prevalent example of this approach is chemoradiotherapy in high-grade gliomas. Although the use of Temodar with concomitant radiation therapy has become the standard of care in adults, there is unclear benefit to this pairing in pediatric patients, and thus there is a need for ongoing studies to investigate new agents to pair with radiation. Administration of chemotherapeutic agents more directly into the CNS through intra-arterial infusions has also been studied as well as chemical or nanoparticle modification.29,34–41
An alternative approach to optimizing delivery of chemotherapy into the CNS is to bypass the BBB by direct administration of chemotherapy, examples of which include intrathecal or intraventricular drug administration, surgical implantation of chemotherapy impregnated devices, and convection enhanced delivery. Intrathecal administration of chemotherapy has become the mainstay of the prevention or treatment of CNS leukemia and lymphomas.42–45 It has also been extensively studied and utilized in the treatment of leptomeningeal disease and for craniospinal prophylaxis especially in younger children with brain tumors.46–51 It is, however, limited in usefulness for bulkier disease due to limited depth of penetration into the brain parenchyma.
Convection-enhanced delivery (CED) is a new approach to enhance the penetration of locally administered chemotherapy. A catheter is placed intraoperatively, usually after surgical re-section of the bulk of the tumor. The addition of convection to simple diffusion has been shown to enhance the penetration of chemotherapeutic agents or toxins into the surrounding tumor and/or brain tissue.40,52–56 Clinical trials utilizing CED in pediatric patients with brainstem and high-grade glioma are ongoing.
Central Nervous System Drug Delivery
Although there are barriers to drug delivery into the CNS, the use of chemotherapy in the treatment of children with brain tumors has grown exponentially over the previous decades. In general, cancer chemotherapeutics achieve their specificity through toxicity to all dividing cells. The necessary result of this approach is a narrow therapeutic range. Drug interactions or changes in dose intensity due to toxicity may affect efficacy; thus a basic understanding and review of the general principles of pharmacology is important for anyone treating children with chemotherapy. Similarly, changes in the pharmacokinetics of a drug may have profound effects on toxicity, and therefore an awareness of potential drug interactions and monitoring of organ function are critical.
In treating pediatric patients, there are several age-related physiological differences that may influence the pharmacokinetics of chemotherapeutics. These differences are most notable in infants, as maturation of organ function is still occurring. Infants in general have decreased clearance and therefore increased exposure to many drugs due to impaired metabolic clearance and immaturity of phase I and phase II metabolizing enzymes, in addition to decreased renal excretion and tubular secretion and decreased protein binding. Beyond infancy, young children may also have increased gastrointestinal (GI) motility, which may affect the absorption of orally administered agents. The volume of distribution of agents may vary significantly over the pediatric age range due to changes in blood volume, extra-cellular and total body water, and body fat composition.57 Each of these variables should be considered, especially when extrapolating treatment regimens and dosing between age groups.58–60 In very young children, or when employing drugs with minimal pediatric experience, involvement of a clinical pharmacologist may be warranted.
Another pharmacokinetic factor that has become increasingly important for clinicians to consider, especially with the increase in the use of oral chemotherapeutic agents, is bioavailability. Although agents that are orally bioavailable often have good CNS penetration, as many of the same transporters affect absorption or efflux at the intestinal lumen are present at the BBB, gastric pH (which may be affected by antacid medications), gastric emptying time, concomitant medications, and administration with food or on an empty stomach may affect the bio-availability of cancer chemotherapeutics. Although many agents are labeled to be taken on an empty stomach, their administration with food may dramatically increase their bioavailability and thus potential toxicity.61 Conversely, interactions that decrease the bioavailability of an agent may impair efficacy. Thus clinicians must be aware of the recommended method of administration as well as potential interactions that may affect bioavailability so that they can appropriately counsel patients on the correct way to take these medications.
Another important drug interaction that must be considered is the potential interaction with antiepileptic medications. Because patients with brain tumors are at increased risk of seizures secondary to their tumor and surgery and may even present with seizure, many are given anticonvulsant medications for seizure prophylaxis. Several of the anticonvulsant medications are strong inducers of the metabolic cytochrome p450 enzymes ( Table 14.2 ). Many commonly used chemotherapies for children with brain tumors such as cyclophosphamide, etoposide, and vinblastine are metabolized by CYP3A4. Therefore, concomitant use of enzyme-inducing anticonvulsants and chemotherapy metabolized by 3A4 will have altered metabolism and exposure and thus altered efficacy or toxicity. Due to the complexity of the pharmacokinetic profiles of many agents, the use of non–enzyme-inducing anticonvulsants should be strongly considered where possible in patients receiving concomitant chemotherapy.
Enzyme-Inducing Anticonvulsant Drugs
Non–Enzyme-Inducing Anticonvulsant Drugs