Cancer Chemotherapy & Biotherapy: Principles & Practices, 4th Edition

Delivering Anticancer Drugs to Brain Tumors

Maciej M. Mrugala

Tracy T. Batchelor

Jeffrey G. Supko

There were an estimated 41,300 new primary brain tumors diagnosed in the United States in 2004.1 Malignant gliomas (anaplastic astrocytoma and glioblastoma) are the most common malignant primary brain tumors and represent the most frequent indication for cytotoxic chemotherapy in neuro-oncology. The goal of adjuvant chemotherapy for malignant glioma is eradication of the residual macroscopic and microscopic tumor felt to be the reason for surgical and radiation failure. Temozolomide, an orally administered alkylating agent, significantly extends progression-free and overall survival when administered concurrently with radiation in patients with newly diagnosed glioblastoma.2 In addition, locally delivered chemotherapy in the form of 1,3-bis(2 chloroethyl)-1-nitrosourea (carmustine; BCNU) polymers also extends survival in patients with malignant glioma when applied at the time of the initial debulking procedure.3However, the survival benefits of adjuvant chemotherapy for patients with malignant gliomas are modest, as demonstrated by an absolute increase in 1-year survival of 6% in one meta-analysis of 12 randomized clinical trials.4

Mechanisms of chemotherapy resistance of brain tumors include factors common to other tumors such as multidrug resistance and increased efficiency of DNA damage-repair systems.5 In addition, treating malignant brain tumors represents a unique challenge for oncologists due to the presence of the blood-brain barrier (BBB), a physiologic impediment between the circulatory system of the brain and that of the body. The accessibility of many anticancer drugs to brain tumors is at least partially constrained by the BBB. Therefore, difficulty in achieving adequate and sufficiently sustained levels of the cytotoxic moiety at the tumor site is a significant factor contributing to the failure of systemic chemotherapy for malignant brain tumors.6, 7 Accordingly, the development of treatment strategies for brain tumors has emphasized techniques that are intended to overcome this barrier and improve drug delivery to these tumors. In addition, the use of multiple ancillary agents in the medical management of brain tumor patients, particularly glucocorticoids and enzyme-inducing antiseizure medications, increases the risk for drug interactions that may impact the efficacy or toxicity of chemotherapy. This chapter reviews the current state of approaches for delivering anticancer drugs to brain tumors, the various techniques that are available for assessing drug distribution to brain tumors, and important pharmacologic interactions that may affect both the accessibility of anticancer drugs to the CNS and the systemic pharmacokinetics of the anticancer agent.


Three main factors influence the extent to which a systemically administered anticancer agent distributes into the brain and brain tumors: (a) the plasma concentration-time profile of the drug; (b) regional blood flow; and (c) transport of the agent through the BBB and blood-tumor barrier (BTB). The two former considerations are common to all solid tumors, whereas the latter is specific to brain tumors.6 Erhlich was the first to propose the concept of the BBB at the beginning of the 20th century. On administering the dye trypan blue to rats by intravenous injection, he observed that all body organs were stained except for the brain and spinal cord.8 The anatomic basis of the BBB was determined three decades ago with the introduction of the electron microscope. It results from a modification of the normal vascular endothelium whereby a sheet of cells is connected by tight junctions on a basement membrane (Fig.21.1). The area of the exchange surface is 12 m2, and the physiologic role of the BBB is assumed to include maintenance of a constant biochemical content of the interstitial milieu and protection of the brain from foreign and undesirable molecules.9 Low hydraulic conductance, low ionic permeability, and high electrical resistance contribute to the very low permeability for hydrophilic nonelectrolytes in the absence of a membrane carrier.10 These properties, together with the lack of intracellular fenestrations and pinocytotic vesicles and the presence of a thicker basal lamina, create a physiologic barrier that is relatively impermeable to many water-soluble compounds.11, 12

Figure 21.1 Schematic representation of the normal intracerebral capillary. Crucial elements for building the blood-brain barrier (basement membrane and endothelial tight junctions) are visualized. (Reproduced with permission from Karen Francis, Johan van Beek, Cecale Canova, et al. Innate Immunity and Brain Inflammation: the key role of complement. Expert Reviews in molecular Medicine pp. 1–19. Cambridge: Cambridge University Press, 2003.)

Some drugs use specific transport mechanisms present in the endothelial cell to traverse the BBB.13 However, most cytotoxic drugs that gain access to the CNS, as exemplified by the chloroethylnitrosourea alkylating agents, cross the BBB by passive diffusion. Aside from pharmacokinetic properties, the main factors that influence the extent to which these compounds distribute into the CNS include lipid solubility, molecular mass, charge, and plasma protein binding. Specifically, small organic compounds with a molecular weight less than 200 that are lipid soluble, neutral at physiologic pH, and are not highly bound to plasma proteins readily cross the BBB.10

A second component of the BBB is the expression of P-glycoprotein on the luminal surface of brain capillary endothelial cells.14 The presence of P-glycoprotein has been implicated in the active efflux of many chemotherapeutic drugs from the brain, including the vinca alkaloids and doxorubicin. Expression of P-glycoprotein has also been reported in malignant gliomas and may serve as another mechanism of chemotherapy resistance.15, 16 Agents that reverse the function of P-glycoprotein, such as verapamil, may increase passage of doxorubicin across the BBB.

The normal physiologic structure of the BBB is disrupted in vasculature within and adjacent to brain tumors. The barrier is usually more permeable in the center of a malignant tumor as opposed to the well-vascularized and infiltrating edge that exhibits variable degrees of BBB integrity.17 Figure 21.2 contrasts the normal BBB with that disrupted by a brain tumor. Vick and colleagues identified junctional clefts in the endothelial cells of capillaries adjacent to brain tumors.18 These clefts correlated with the density of infiltrating tumor cells and were present in brain capillaries not in direct contact with the tumor. Evidence also exists that the microvasculature of these tumors lacks the properties of a normal BBB and has greater permeability as a result. Morphologic studies have demonstrated that the BTB differs anatomically from the normal BBB, with open tight junctions, gap junctions, fenestrations, and numerous intracellular vesicles.6, 19 The increased permeability of these blood vessels forms the basis of contrast enhancement of brain tumors on CT and MRI scans. Iodinated, water-soluble contrast agents do not penetrate areas of the brain with an intact BBB but are able to penetrate brain tumors.20 These alterations in permeability are highly variable between tumors and within the same tumor.21 For example, low-grade gliomas and proliferating edges of malignant gliomas seem to have a normally functioning, selective BBB and consequently do not typically show contrast enhancement on CT and MRI studies. A large variation in the enhancement patterns of malignant brain tumors on CT scans is common.22, 23 Approximately 30% of patients with a type of malignant glioma, anaplastic astrocytoma, are reported to have nonenhancing lesions.22 Finally, positron emission tomography (PET) studies have shown that alterations in permeability usually occur in the central part of large tumors, whereas the periphery is intact.24 The presence of a selective, normal BBB near the proliferating edge of a brain tumor may result in variable delivery of water-soluble drugs in this region and may contribute significantly to the high local failure rate of conventional anticancer drugs.

Figure 21.2 Heterogenecity of the blood-brain barrier. Contrast-enhanced CT scans of the basal ganglia of a 61-year-old woman with a primary CNS lymphoma, indicating that the permeability of the blood-tumor barrier is inconsistent for a given patient or even a given tumor nodule. A. CT scan demonstrating a bright, uniformly enhancing lesion in the right basal ganglia. The surrounding hypodense signal in the brain tissue around the tumor (arrowheads) should be noted. B. CT scan obtained after contrast agent administration. Contrast material was administered immediately after osmotic BBBD and CT scans were obtained 30 minutes after the first BBBD treatment in order to confirm and assess the grade of BBBD. The patient underwent right internal carotid artery disruption in the anterior and middle cerebral artery distributions (arrows). Opening of the brain tissue around the tumor in the area of the peritumoral hypodense signal evident in the CT scan in A should be noted. C. CT scan obtained after BBBD in a patient with a right hemiparesis that was unexplained, because the only visible tumor was in the right cerebrum (A). BBBD the day after the CT scan in (A) extended into the posterior circulation via the posterior communicating artery. A left-side brainstem lesion not apparent in pre-BBBD imaging studies was noted. The right hemiparesis was thus attributable to a brainstem tumor (arrows) on the left that was not apparent on pre-BBBD MRI scans (intact BBB and no edema). (Reproduced with permission from Neuwelt EA. Mechanisms of disease: the blood-brain barrier Neurosurgery 2004;54:131–140.)

The central role hypothesized for the BBB in the resistance of brain tumors to chemotherapy has been questioned by some authorities. It has been suggested that, because the BBB adjacent to tumors and the BTB lack the normal properties of an intact BBB, drug delivery to these areas should not be compromised.18Indirect support for this argument comes from the observation that brain tumors occasionally respond to cytotoxic drugs that would not be expected to cross the BBB due to their physicochemical properties. A more consistent observation, however, has been that the most effective classes of antineoplastic drugs against malignant brain tumors are lipid-soluble molecules that can easily penetrate an intact BBB. Moreover, in addition to the existence of normal BBB at the proliferating edge of brain tumors, PET studies have demonstrated that, whereas the BBB and BTB may be abnormal at the time of diagnosis, these structures may become normalized on subsequent treatment.25 These latter observations support a pivotal role for the BBB and BTB in the resistance of brain tumors to systemically administered chemotherapy.


Following the administration of an anticancer drug by the intravenous or oral routes, the BBB can effectively impede the distribution of drug molecules from systemic circulation into the CNS. Consequently, considerable effort has been expended to develop drug delivery strategies that either entirely circumvent the BBB or modulate the permeability of the barrier to enhance the extent of drug distribution into the brain from systemic circulation. These techniques include intra-arterial administration, BBB disruption with hyperosmolar solutions26 or biomolecules, high-dose chemotherapy, intrathecal injection, and local delivery by direct intratumoral injection of free drug or implantation of drug embedded in a controlled-release biodegradable delivery system. Even when drugs have crossed the BBB, however, their migration to tumor cells may be hindered by increased intercapillary distances, greater interstitial pressure, lower microvascular pressure, and the sink effect exerted by normal brain tissue.27

Intra-arterial Administration

The theoretical advantage for delivering anticancer drugs by the intra-arterial route is related to the ratio of systemic to regional blood flow. In comparison to intravenous injection, a considerably higher local drug concentration can be achieved with intra-arterial injection, thereby increasing the amount of the agent driven across the BBB. This has been confirmed experimentally, including in a report of a two- to threefold increase in tumor concentration of cisplatin and a chloroethylnitrosourea, respectively, in the brains of animals after intra-arterial administration as compared to intravenous administration.28 With this technique, sufficient local drug concentrations can be achieved with smaller than conventional doses, so that systemic side effects are minimized.29 The pharmacokinetic advantages of intra-arterial administration occur only during the first passage through the CNS, because the drug then enters venous circulation, and the plasma profile is indistinguishable from that afforded by intravenous administration.

This approach has been clinically evaluated in various settings, including neoadjuvant and adjuvant therapy and recurrent malignant glioma.30, 31, 32 Thus far, clinical trials of intra-arterial chemotherapy for malignant gliomas have not demonstrated improvement in survival over conventional intravenous therapy. A phase III study involving 315 patients with malignant gliomas failed to show any advantages of adjuvant intra-arterial BCNU over intravenous infusion of the same drug.33 Moreover, subjects in the intra-arterial BCNU arm of this trial experienced significant treatment-related toxicities, with 10% developing leukoencephalopathy and 15% developing ipsilateral blindness. Another randomized clinical trial compared intra-arterial and intravenous ACNU (nimustine). There was no significant difference in the progression-free survival and overall survival in each treatment arm. However, toxicity associated with intra-arterial ACNU was modest. No cases of leukoencephalopathy and only one case of transient visual impairment were reported.34

Potential disadvantages of the intra-arterial route include local complications related to catheterization (thrombosis, bleeding, infection) and neurological sequela, including orbital and cranial pain, retinal toxicity, leukoencephalopathy, or cortical necrosis.30, 33, 35 In addition, prodrugs requiring hepatic activation, such as cyclophosphamide, procarbazine, and irinotecan, are not suitable for use by the intra-arterial route One factor that partly explains the unique toxicities associated with intra-arterial administration is the “streaming” effect.36 Infusion of drug into the high-pressure, rapidly moving arterial bloodstream results in incomplete mixing of the drug and plasma and great variability in the amount of drug reaching different regions of the vascular territory. Depending on the characteristics of the distribution pattern, higher concentrations of drug might be achieved in normal brain, whereas lower amounts reach the tumor. Different strategies have been attempted to minimize the streaming effect, including rapid infusion,36 superselective cannulation of the feeding artery,35, 37 diastole-phased pulsatile infusion,38, 39 and local blood flow adjusted dosage.37 All of these techniques were combined in a phase I trial involving 21 brain tumor patients treated with intra-arterial carboplatin. The neurologic side effects were minor, and a twofold escalation of the dose beyond the conventional intra-arterial dose was achieved, with promising results.37 Despite its serious limitations, there have been a few reports of promising results in patients receiving intra-arterial therapy for primary brain tumors. Intra-arterial carboplatin and etoposide were demonstrated to be safe and useful in the treatment of progressive optic-hypothalamic gliomas in children.40 The safety and efficacy of intra-arterial chemotherapy with multiple agents in conjunction with osmotic disruption of the BBB were established.41 It seems that intra-arterial chemotherapy, with all its current limitations, is slowly moving forward and may achieve wider applications. Perfection of the delivery techniques and development of newer, less toxic compounds should increase its efficacy and safety.

Blood-Brain Barrier Disruption

Blood-Brain Barrier (BBB) disruption involves the use of hyperosmolar solutions or biomolecules to increase the permeability of the BBB and improve drug delivery to brain tumors. The specific mechanisms underlying osmotic opening of the BBB and BTB are not entirely understood. Preliminary explanations emphasized endothelial cell shrinkage and resultant separation of tight junctions upon exposure to a hyperosmotic environment.42 In addition to cellular shrinkage, osmotic stress releases biologically active compounds from endothelial cells, including serine proteases, that could potentially degrade the collagen matrix of the endothelial basement membrane. Finally, cellular shrinkage may also trigger second messenger signals and calcium influx, which could affect the integrity of tight junctions.9

Methods for disrupting the BBB involving both intravenous and intra-arterial administration have been developed for use in brain tumor patients.42 The results of nonrandomized studies have been encouraging for certain brain tumor subtypes, especially primary CNS lymphoma. Potential advantages of this method include increased tumor delivery of drug and lack of systemic toxicity from cytotoxic chemotherapy. Another possible advantage of this technique is avoiding the sink effect seen with other procedures used for delivering chemotherapy to brain tumors. The sink effect refers to the selective achievement of higher concentrations of a drug in areas of disrupted BBB in the tumor than in the rest of the brain. As a result of this concentration gradient, the drug rapidly diffuses out of the tumor into the surrounding brain and compromises tumor exposure time. Because BBB disruption theoretically affects the endothelium of both normal brain and brain tumor, a nonselective increase in drug delivery into both areas occurs, and no concentration gradient is established. Despite these potential advantages, the technique of BBB disruption is complex and requires transfemoral angiography and general anesthesia. Moreover, an attendant risk of stroke as well as a high frequency of seizures are associated with this method. These factors have limited the application of this technique.43, 44

Tumor location and vascular supply determine the arterial circulation that is catheterized and infused. Most commonly, one major artery (left carotid, right carotid, left or right vertebral) is cannulated and treated. Some have advocated that documentation of BBB disruption with iodinated contrast agents be obtained before chemotherapy administration. Given the technical requirements of this procedure, it has not been widely adopted, and no definitive conclusions about the efficacy of the technique can be derived from the results of clinical trials that have been published to date. Other, less invasive techniques of BBB disruption are being investigated. It was recently shown that focused ultrasound exposure, when applied in the presence of preformed gas bubbles, can cause MRI-proven reversible opening of BBB in rabbits.45 There is also interest in evaluating the use of biological agents to increase permeability of the BBB. Experimental data have demonstrated the effectivness of several such vasoactive compounds, including histamine, leukotriene C4, beta-interferon, tumor necrosis factor-alpha, and bradykinin. The bradykinin analog RMP-7 (Cereport) selectively increases delivery of radiolabeled carboplatin to brain tumor in animal xenograft models and improves survival. However, a recent randomized double-blind, placebo controlled phase II study showed that RMP-7 does not improve efficacy of carboplatin in patients with recurrent malignant glioma.46

High-Dose Chemotherapy

Considering that the BBB is a major factor in brain tumor resistance to chemotherapy and that diffusion across this barrier depends on the concentration-time profile of the free fraction of drug (i.e., drug that is not bound to plasma proteins), the assumption has been that increasing the administered dose would drive more drug across the BBB.6 The rationale for high-dose chemotherapy (HDCT) was derived from the relatively linear in vitro dose-response curve exhibited by the classic alkylating agents and the assumption that intrinsic cellular resistance could be overcome by increasing the dose. In the context of treating brain tumors, the argument has been made that HDCT could overcome the previously mentioned sink effect and provide higher drug concentrations in the tumor for sustained periods. A number of phase I and II studies have been undertaken to evaluate this approach.47 Despite the theoretical advantages, HDCT for recurrent malignant glioma has not made a significant impact on patient survival, although cases of long-term survival have been observed anecdotally. Among patients with newly diagnosed tumors, the median survival achieved using HDCT with bone marrow or stem cell rescue is comparable to that with conventional-dose chemotherapy (12 to 26 months versus 10 to 12 months, respectively). Treatment-related morbidity and mortality have been high, however, with a mortality rate as great as 27% in one early study. More recently, HDCT with autologous stem cell rescue was successfully used in treatment of medulloblastoma and malignant astrocytic tumors in children.48 Safety and efficacy of this modality was also addressed by investigators in the treatment of recurrent CNS germinomas and malignant astrocytomas in adults.49, 50 Prolonged tumor control was achieved when HDCT in conjunction with stem cell rescue was used in the treatment of newly diagnosed anaplastic oligodendroglioma.51 The strategy of HDCT followed by stem cell rescue may become an effective treatment strategy for potentially chemosensitive brain tumors such as anaplastic oligodendroglioma and primary CNS lymphoma.52

Intrathecal Administration

Intrathecal chemotherapy involves the direct injection of drug into the CSF and is an obvious way to bypass the BBB. It is accomplished by injecting drug into the lumbar subarachnoid space, the cerebral ventricles, or the basal cisterns, with or without the use of catheters, pumps, or ventricular reservoirs.53 The rationale is that the cells lining the fluid spaces of the brain are permeable, which results in a free exchange of molecules from extracellular fluid (ECF) to CSF and vice versa. Relatively small doses of a drug given by intrathecal injection can achieve high local concentrations due to the low volume of CSF (approximately 150 mL), minimizing systemic toxicity. Furthermore, because of the intrinsically low levels of enzymes in the CSF, some potentially useful agents that are subject to rapid metabolism in blood remain in the active form in the CNS for a longer periods of time. The three drugs used most commonly in this manner are methotrexate sodium, cytosine arabinoside (ara-C), and thiotepa (thiotriethylene phosphoramide). These agents have been used mainly for the treatment of leptomeningeal metastases from systemic cancer.54

Figure 21.3 The subarachnoid spaces and cisterns of the brain and spinal cord. Schematic representation of cerebrospinal fluid pathways (arrows). (Reprinted with permission from Fix J. High-Yield Neuroanatomy. 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2000.)

Intrathecal drug administration has several disadvantages, including the necessity to establish access to the CSF compartment. A ventricular access device is usually implanted in the frontal horn of the lateral ventricle to facilitate the administration of drug and minimize patient discomfort. This entails a small surgical risk of hemorrhage and infection. Moreover, the catheter may malfunction over time and require replacement.55 Intrathecal drug administration also has numerous pharmacokinetic limitations. Among these, the drug must overcome bulk flow of CSF to penetrate the cisterns and ventricles (Fig. 21.3). In addition, the flow of interstitial fluid produced by brain cells and microvessels from ECF to CSF counteracts the diffusion of drug from CSF to ECF. Estimates are that CSF is completely renewed every 6 to 8 hours; thus, the concentration of a drug injected into the CSF decreases continuously as a consequence of this process, which can only be overcome by a continuous infusion or sustained-release system to maintain a clinically relevant concentration. Development of a liposomal form of ara-C allows sustained release of this drug into CSF and increases the effective half-life of the agent in the CSF by almost 50-fold.56
Another pharmacologic disadvantage of the intrathecal route is the fact that production of CSF by the choroid plexus and its elimination into the venous circulation may be altered by the tumor itself, which disturbs bulk flow and modifies drug distribution and diffusion. For example, the clearance of methotrexate from CSF is decreased in the presence of leukemic meningitis.57, 58 Moreover, diffusion in the ventricular space is heterogeneous59 and may result in uncertain and potentially toxic local concentrations in CSF, even if continuous-infusion devices are used.53 Finally, and most important, brain tumors are often in locations not adjacent to the ventricular system and may require diffusion of drug from CSF to tumor over a distance of several centimeters, which impairs the ability to achieve cytotoxic concentrations of the drug at the site of the tumor. In the case of methotrexate, the concentration of drug has been calculated to be no more than 0.1% of the CSF concentration 1 cm from the ependymal edge 48 hours after intrathecal administration.60 However, the relationship is quite complex. Whereas compounds with greater lipophilicity will access the ECF more effectively, they will also be subject to a higher rate of removal by the vascular and cellular compartments, thereby limiting the extent of drug penetration into brain tissue.61 Therefore, at this point in time, intrathecal chemotherapy is not feasible for brain tumors and is restricted to the treatment and prevention of leptomeningeal metastases.62

Intratumoral Administration

The simplest and most direct way to guarantee that a cytotoxic drug reaches its target is to deliver it directly into the tumor or into the cavity left after tumor resection. As with intrathecal injection, this bypasses most of the previously mentioned obstacles pertaining to systemic drug administration for treating brain tumors. Systemic toxicity may also be reduced because substantially lower doses of drug may be given, and only a relatively small amount of drug distributes from the CNS to the bloodstream. A particularly attractive advantage of this strategy is that anticancer drugs that are normally impeded by the BBB may be used. Conceptually, the low permeability of the BBB to such compounds should promote their retention within the CNS by inhibiting distribution into the bloodstream. Due to the high local drug concentrations that can be achieved, better distribution of drug may be provided within the tumor by diffusion and convection driven by the hydrostatic pressure of the tumor. The two techniques that have been most commonly used for directly introducing chemotherapeutic agents into brain tumors are (a) parenteral delivery as either a bolus injection or continuous infusion through a cannula and (b) implantation of drug embedded in a slow-release carrier system.

The feasibility of intratumoral infusion has been demonstrated in a number of clinical trials involving approximately 10 different anticancer agents.63 These studies, however, have not shown a clear survival advantage or direct evidence of increased drug delivery within the tumor. Furthermore, toxicity has been observed with this technique, including nervous system injury and infection.63 Even if intratumoral infusion does avoid some obstacles to drug delivery, it does not circumvent the sink effect or problems associated with drug stability. Indeed, drug molecules released into the ECF must penetrate the brain interstitial tissue to reach tumor cells.27 Before reaching its target, the compound could be inactivated by binding to normal tissue, metabolism, chemical degradation, or elimination by the microvascular circulation.64 Finally, obstruction of the catheter by tissue debris can occur.63, 64

Controlled-release methods using polymer, microsphere, and liposomal carriers have been studied extensively in vitro and in vivo.7 The goal of this strategy is to provide constant delivery of a cytotoxic drug into the tumor using a matrix that also protects the unreleased drug from hydrolysis and metabolism. Use of a solid polymeric matrix to facilitate the delivery of chemotherapeutic agents directly to brain tumors has several potential advantages. Biodegradable carriers have been developed that are unaffected by interstitial pH and provide near zero-order release of drug, with minimal inflammatory response.65 Potential disadvantages include the fact that drug release cannot be controlled once the device has been implanted without physically removing it. Other potential problems include unpredictable diffusion, stability of the device and drug in the aqueous milieu, and the possibility that the polymer may not release the drug as intended.

A biodegradable polyanhydride solid matrix, poly [bis(p-carboxyphenoxy)propane-sebacic acid] or p(CPP-SA), has been developed that releases drug by a combination of diffusion and hydrolytic polymer degradation.66 Preclinical studies have demonstrated that this system is biocompatible and results in reproducible and sustained continuous release of BCNU. More than 300 patients with recurrent67, 68 and newly diagnosed3, 69, 70 malignant gliomas have been treated with the BCNU polymer in phase I–II clinical trials and phase III placebo-controlled studies. A phase III study in patients with recurrent malignant glioma demonstrated that intratumoral implantation of a 3.85% BCNU polymer was safe and resulted in minimal systemic side effects from BCNU. Median survival was significantly longer in subjects with glioblastoma who received the active polymer than in those who did not, even after adjustment for known prognostic factors.67 A phase III randomized placebo controlled trial examining BCNU polymer application in glioblastoma patients at the time of primary surgical resection also showed survival benefit.3 A phase I study designed to increase the amount of BCNU in the polymer (up to 20%) demonstrated that at the highest doses BCNU plasma levels were significantly (500-fold) lower than those associated with systemic BCNU toxicity. As a result, no BCNU-related systemic toxicity was seen.71 However, risk of local neurotoxicity may be increased at higher BCNU concentrations in the polymer. Other studies of this delivery strategy are assessing use of the BCNU polymer in combination with systemic chemotherapy and incorporating other anticancer drugs into the polymer.72 A clinical trial to evaluate intratumoral implantation of 5-fluorouracil–releasing microspheres has been initiated.73

Nonchemotherapeutic approaches involving intratumoral administration are also being investigated. A fusion protein consisting of interleukin 13 (IL-13) linked to a mutated form of Pseudomonas exotoxin (IL-13 PE38QQR) that is administered by convection-enhanced delivery (see below) has entered phase I/II trials.74 In addition, intratumoral administration of chimeric proteins of transforming growth factor (TGF-α) and mutated Pseudomonas exotoxin PE-38 (TP-38) directed against the epidermal growth factor (EGFR) is being studied.73

Convection-Enhanced Delivery

Experimental evidence has demonstrated that properties of the brain parenchyma may impede delivery of drugs to the site of a brain tumor. Therefore, the BBB may not be the only obstacle that must be overcome for successful delivery of a cytotoxic drug to a brain tumor. Diffusion barriers intrinsic to brain tissue may also be important in limiting drug delivery to tumors. The hydrostatic pressure of brain tissue and the solubility of the drug are important factors that determine the diffusion of drug into surrounding tissue. Convection-enhanced delivery (CED) is a pressure gradient–dependent method developed to overcome these potential barriers; it consists of a direct infusion of drug solution through a catheter surgically implanted in the brain tumor.75 Experimental studies of CED have demonstrated that drug delivery with this method is dependent on the anatomic site of the catheter. Infusion into gray matter results in spherical distribution of the drug, whereas infusion into white matter results in distribution along white matter fiber pathways. Therefore, the specific anatomic location of the brain tumor may be an important determinant of drug delivery.

Studies of the delivery of ara-C to rat brain after intravenous, intrathecal, and intraventricular administration and CED have been conducted.76 Using quantitative autoradiographic analysis, it was demonstrated that drug concentration in brain tissue after CED was 4,000-fold higher than after intravenous administration. Moreover, the volume of distribution was 10-fold higher after CED than after intrathecal or intraventricular administration. Experience with the use of CED in the treatment of patients with brain tumors has been limited. In a phase I study involving 18 patients with recurrent malignant gliomas, a high-flow interstitial microinfusion of a conjugated form of diphtheria toxin was conducted.77 In 9 of 15 evaluable patients, at least a 50% regression of tumor was apparent on MRI, and two complete responses were observed. The treatment was well tolerated, and no treatment-related deaths or systemic toxicity occurred. The dose-limiting toxicity was local brain injury, which may have been the result of endothelial damage of cerebral capillaries.


The clinical effectiveness of chemotherapy ultimately depends upon exposing tumor cells to adequate concentrations of the biologically active form of anticancer drugs for a sufficient duration of time. Distribution of drug from the administration site to the tumor and its subsequent elimination from the body are dependent upon the physicochemical properties of the drug and numerous physiological factors. Penetration of the BBB or BTB is an additional complexity in the use of parenterally or orally adminstered anticancer agents against tumors residing within the CNS, one that is not a consideration in the treatment of hematological malignancies or solid tumors.6, 27, 78 Characterizing the exposure of brain tumors to chemotherapeutic agents presents an extremely challenging problem. As is the case with any organ or tissue, the time course of the concentration of a drug or active metabolite within a tumor cannot be discerned from experimental data limited to measurements made in plasma, serum, or whole blood. Although some temporal relationship undoubtedly exists between drug concentrations in plasma and those in tumor, elucidating the tumor concentration-time profile necessarily requires measuring drug levels within the tumor itself.

Determining whether adequate concentrations of the active form of a drug reach the target tissue is extremely important in the context of phase I trials to evaluate the efficacy of new anticancer drugs for treating brain tumors. Because objective antitumor responses occur infrequently in phase I studies, the availability of data regarding the extent to which a chemotherapeutic agent reaches a brain tumor would provide a rational basis for selecting drugs that warrant further clinical evaluation. As described in this section, the principal techniques that are less invasive and potentially more informative than tissue biopsy studies for assessing the pharmacokinetics of anticancer drugs in the CNS and brain tumors include CSF sampling, microdialysis, and noninvasive imaging. Although no single method can be uniformly applied for monitoring drugs in human tissues in vivo, these techniques are nevertheless becoming increasingly important to the clinical development of anticancer drugs for the treatment of brain tumors.

Direct Measurment in Tumor Tissue

The traditional method for evaluating drug distribution to a solid tumor by directly measuring its concentration in tissue has numerous deficiencies. Subjecting brain cancer patients to the risks of an intracranial surgical procedure that may have little or no direct benefit to the treatment of their disease raises ethical concerns. It may be possible to circumvent this problem by obtaining tissue when a biopsy was diagnostically indicated or during a necessary tumor debulking procedure. Nevertheless, measuring the concentration of drug in a single biopsy specimen provides very limited information unless the tissue is obtained while drug is being given in a manner that provides continuous systemic exposure to the agent. Otherwise, the most appropriate time to obtain a single biopsy specimen relative to drug administration is speculative at best, as the presence or absence of a measurable drug concentration has little interpretive value. Acquiring serial tumor specimens from the same patient presents even greater practical constraints than conducting a single biopsy, and the effect of prior procedures on altering the transport of drug to and from the tumor represents a significant confounding factor. Conceivably, as part of a phase II study in which a moderately sized cohort of patients with comparable disease characteristics are treated with the same dose of drug, single biopsies of the tumor and adjacent peritumoral tissue could be performed in different patients over a range of times relative to drug administration, allowing a composite or pooled tissue concentration-time course to be constructed.

Cerebrospinal Fluid

Pharmacologic studies of anticancer agents directed against brain tumors often include the determination of drug or drug metabolite concentrations in the CSF as a surrogate for tumor levels and as a measure of drug delivery beyond the BBB. An understanding of the composition and normal physiology of CSF is important to discern the significance of drug level monitoring in this compartment. The most distinctive difference between CSF and plasma is the substantially lower concentration of proteins in CSF. Because of this, the total concentration of compounds that are poorly soluble in water or that bind avidly to proteins would be expected to be lower in CSF than in plasma or brain tissue, although the free concentration may be increased. In addition, CSF is slightly more acidic than plasma (pH of 7.32 versus 7.40), and this differential could conceivably influence the transport and retention of a drug in CSF, as well as its chemical stability relative to that in plasma, for compounds that have a functional group with a pKa in the 7 to 8 range.

Fig. 21.3 depicts the normal process involved in CSF formation and flow. The volume of CSF contained within the ventricles, cisterns, and subarachnoid space of a normal adult is approximately 150 mL.79 Approximately 500 mL of CSF is produced every 24 hours, predominantly by the choroid plexus within the cerebral ventricles; therefore, the entire CSF volume is replenished three times during the course of a day. CSF formed in the lateral ventricles flows into the third ventricle and then into the fourth ventricle. Upon exiting from the fourth ventricle, it passes into the basal cisterns and the cerebral and spinal subarachnoid spaces, descending through the posterior aspect of the spinal cord and returning through the anterior aspect. Ascending CSF passes over the cerebral hemispheres toward the major dural sinuses, where absorption of CSF into the venous system occurs at the arachnoid villi. The presence of a brain tumor can significantly diminish both the formation and flow of CSF.80

Concentration gradients between the ventricular and lumbar regions exist for endogenous constituents of CSF as well as for xenobiotics that have gained access to the CSF. The concentration of systemically administered drugs is generally higher in CSF collected from the ventricles than from the lumbar region, as drug distribution in the CNS follows CSF flow.81 Because drug levels are often determined in CSF acquired by lumbar puncture, it should be recognized that this may significantly underestimate the concentrations in the ventricular region. Similarly, drug administered directly into lumbar CSF is poorly distributed to the ventricles.59, 82 Although patients cannot be subjected to frequently repeated lumbar punctures, ventricular access devices such as the Ommaya reservoir may be used to facilitate the serial acquisition of CSF specimens for drug level monitoring in the brain. Therefore, the specific space from which CSF samples were collected must be taken into account whenever drug concentrations reported in different studies are compared. Furthermore, in the process of collecting CSF specimens, the fluid balance or bulk volume of CSF could be significantly altered, which affects pressure equilibrium and flow between the various CSF compartments as well as the concentration gradients between CSF and plasma.

The presence of drug in the CSF represents a strong, but not definitive, indication that the compound has gained access to brain tissue, inclusive of tumors. Some level of uncertainty exists because the vascular supply to the choroid plexus, hypophysis, and pineal gland is not protected by the BBB. Although comprising a comparably small exchange surface, these do provide a direct pathway for the transport of compounds into the CSF. To further complicate matters, the absence of measurable drug or metabolite levels in the CSF does not absolutely imply that the agent has failed to reach a brain tumor. Conceivably, the majority of a compound reaching brain tissue could be trapped in the intracellular space, extensively bound to tissue protein, subject to chemical or enzymatic conversion to unknown products, or effluxed back into the bloodstream before migrating from interstitial spaces to the CSF. Accordingly, many drugs have been found to achieve higher concentrations in a brain tumor than would have been predicted from plasma and CSF data.83, 84 Despite the limitations of CSF as an indirect measure of drug delivery to a brain tumor, it is an accessible compartment, and CSF sampling remains an important method for screening drug access to the CNS.


Inadequate transport of drug from systemic circulation to the interstitial space surrounding tumor cells is considered one of the primary reasons for the failure of chemotherapy for malignant gliomas.27 Microdialysis is a technique that enables the concentration of a compound with appropriate physicochemical characteristics to be continuously measured in ECF within a tumor or normal tissues in a living subject.85 Commercially available microdialysis catheters consist of a chamber that is less than 1 mm in diameter fitted with a semipermeable membrane that has a molecular weight cutoff ranging from 20 to 100 kd and into which two sections of microcatheter are fused. The device is stereotactically implanted such that the membrane resides within the desired area of the brain or tumor and the microcatheters are externalized. One of the microcatheters is connected to a syringe pump containing an appropriate perfusion fluid, and dialysate is collected from the other microcatheter. A sterile solution that approximates the composition of CSF, typically a Ringer's solution, is used as the perfusion fluid and delivered at flow rates in the microliter-per-minute range. In theory, microdialysis mimics the passive function of a capillary blood vessel.85-87 Water, inorganic ions, and small hydrophilic organic molecules freely diffuse across the membrane of the probe, which is impermeable to proteins and protein-bound compounds. Lipophilic organic compounds are poorly recovered.

Considerable experience has been gained with the use of microdialysis, in both animal models and human subjects, since the technique was first introduced some 3 decades ago.88 The greatest advantages of the technique for pharmacodynamic and pharmacokinetic studies is the facilitation of direct, serial sampling of ECF from a highly localized area of intact tissue.78, 89 Microdialysis has been thoroughly evaluated in head injury patients for monitoring lactic acid, glucose, glutamic acid, γ-aminobutyric acid, oxygen partial pressure, and pH, both for diagnostic intent and for assessing the effect of therapeutic interventions.78, 90-92 Insertion and removal of the dialysis probe results in minimal injury to normal brain tissue, alteration of fluid balance, or disruption of the BBB.78, 93, 94 Although it can be performed safely, microdialysis is nevertheless an invasive surgical procedure with a small risk of bleeding at the insertion site and of discomfort for the patient.

Due to the low flow rate at which the perfusion solution is delivered, the perfusate must generally be collected over intervals ranging from 5 to 30 minutes in order to obtain a sufficient volume for chemical analysis. Consequently, microdialysis may not be suitable for studying systems in which the concentration of the compound of interest is changing rapidly relative to this time frame. The perfusate obtained by microdialysis can be analyzed directly by methods such as reversed-phase high-performance liquid chromatography, liquid chromatography/mass spectrometry, or capillary electrophoresis without the need for any preliminary sample preparation to remove macromolecules, as is generally required for assays performed on plasma.78, 95, 96 Due to the dynamic nature of the system, in which the concentration of compounds in the ECF may be constantly changing while the perfusion fluid within the dialysis probe is continually flowing, steady-state conditions between the sampled fluid and dialysis fluid are not achieved. Nevertheless, it has been conclusively established that the in vivo recovery of an analyte, defined as its concentration in the dialysate relative to that in the sampled fluid, is independent of the concentration in the sampled fluid. Because the drug concentration in the dialysate is generally much lower than in ECF, the diluting effect of the technique demands a very sensitive analytical method to enable the detection of low concentrations of an analyte in very small sample volumes.78

The initial preclinical application of the technique to monitor the distribution of an anticancer drug to a brain tumor was described by de Lange, who measured the concentration of methotrexate in a rodent brain tumor model.97 The results obtained with microdialysis were comparable to those in previous studies of methotrexate distribution to tumors based on the classic methods of autoradiography and tissue excision.98 Subsequently, the technique has been used to define the concentration-time profile of camptothecin and topotecan in rodent brain tumor models99, 100 as well as of other antineoplastic agents in various preclinical tumor models.85, 86, 101, 102 Despite the demonstrated potential offered by the technique, there are still no published reports in which microdialysis has been used to study the intratumoral disposition of an anticancer drug in brain cancer patients. Publications documenting the use of the technique in patients with non-CNS solid tumors have solely described the monitoring of carboplatin levels in melanomas103 and the delivery of 5-fluorouracil to breast tumors.89 In the latter study, a relationship between the area under the interstitial concentration-time curve of 5-fluorouracil and therapeutic response was established. However, the feasibility of using microdialysis in the reverse application, for intratumoral drug delivery, has been demonstrated in a small cohort of glioma patients.104

Noninvasive Imaging

With continual improvements in spatial resolution, temporal resolution, and sensitivity, imaging techniques such as magnetic resonance spectroscopy (MRS) and PET offer the ability to noninvasively monitor the concentration of a drug or its metabolites within brain tumors and surrounding normal tissue. MRS involves the application of a strong magnetic field to the brain, which induces the absorption of energy in the radiofrequency range by atomic nuclei with appropriate spin characteristics. The frequencies of energy required to effect a transition between nuclear spin states are characteristic for each different nucleus and can be readily distinguished by a spectrometer. The isotopes of atoms amenable to detection by MRS that are of greatest interest with regard to drug level monitoring include 1H, 2H (deuterium), 11B, 13C,15N, 19F, and 31P. These are stable isotopes, although not always the most abundant isotopic form of an atom (e.g., 13C). The spatial location and connectivity of atoms in the immediate vicinity of any given nucleus within a molecule influence the effective magnetic field experienced by the nucleus; this results in small but distinctive shifts in the resonance energy that convey structural information. The physical configuration of the instrumentation used for in vivo MRS results in a very substantial reduction in the resolution that can be achieved with analytical spectrometers. Whereas in vivo MRS can readily distinguish different atomic nuclei, the amount of detail pertaining to the type of molecule or functional group to which similar nuclei are bound is limited.

The intense signals associated with 1H and 31P, which result from the ubiquitous presence of these nuclei in bio-organic molecules, are used advantageously for diagnostic MRI. However, 1H and 31P MRS cannot generally be used to monitor xenobiotics in vivo, due to the intensity of the natural background signals, unless a compound has nuclei with chemical shifts that differ substantially from the resonance frequencies arising from endogenous molecules. Examples of the latter situation include the detection of iproplatin by 1H MRS and ifosfamide by 31P MRS.105, 106 Aside from these considerations, the relatively poor sensitivity of MRS is perhaps the single factor that most severely limits its utilization for in vivo drug level monitoring. A molecule must generally be present at concentrations in the low millimolar range to produce a detectable signal. Relatively few cytotoxic drugs achieve these concentrations in plasma or tissues. Furthermore, the anticancer agents that are currently being advanced into clinical trials tend to be increasingly potent, with MTDs that provide peak plasma concentrations in the nanomolar to low micromolar range in patients. The detection limits of MRS can be improved to some degree by increasing the data acquisition time. This sacrifices temporal resolution, however, which may be extremely important for pharmacokinetic studies. MRS does not permit the determination of the absolute concentration of a drug in vivo, although changes in relative concentration during the course of a single experiment can be readily followed. Another important consideration in using MRS to study the time course of a drug in brain tumors is that the spatial resolution is not particularly good, being approximately 1 cm2 with current instrumentation.107

Despite these limitations, MRS has proven to be extremely valuable for studying the tissue pharmacokinetics of some fluorinated drugs, such as 5-fluorouracil.107-110 The 19F nucleus represents an ideal object for in vivo MRS studies. Spectra acquired from MRS scans of the human body show no background signals in the region where 19F resonates due to the lack of endogenous fluorinated organic compounds. Moreover, the nuclear spin characteristics of 19F provide an excellent signal.107 An MRS study involving 103 patients with extraneural malignancies demonstrated that therapeutic response to 5-fluorouracil was significantly correlated with the half-life of the drug within tumors.107, 109, 111 Gemcitabine hydrochloride is another fluorinated anticancer agent for which 19F MRS has been successfully used for in vivo pharmacokinetic studies.112, 113 Although the feasibility of the approach has been demonstrated, very little clinical experience has been gained with the use of MRS for studying the distribution of anticancer drugs to brain tumors.

Dynamic PET imaging is an established technique for defining the time course of radiolabeled anticancer drugs within brain tumors and surrounding normal tissue in patients. In comparison with MRS, PET enables radiolabeled compounds to be detected with 106 to 109 times greater sensitivity and superior temporal resolution due to shorter data acquisition time.114 Furthermore, whereas MRS is effectively limited to monitoring the relative change in concentration of a compound during the course of a study, PET provides a quantitative measurement of radioactivity. In addition, excellent spatial resolution, on the order of 6 mm or better, can be achieved with the current generation of detectors in whole-body PET scanners.115

The ability to use this methodology ultimately depends on developing a suitable procedure to introduce a positron-emitting radionuclide into the drug molecule. The fact that substituting a stable atom with a radioisotope of the same element does not affect the physicochemical, pharmacokinetic, or biologic properties of a drug has been well established. The radionuclides that have been most commonly used for PET pharmacokinetic studies are 11C, 18F, and 13N.115 Because these radionuclides have very short half-lives, namely 10 minutes for 13N, 20 minutes for 11C, and 110 minutes for 18F, a cyclotron and remote-controlled radiochemical synthesis facility are required for on-site generation of the radionuclide and immediate preparation of the labeled drug.115, 116As a consequence, the technique is expensive, and relatively few institutions in the United States have assembled the physical facilities required to undertake PET pharmacokinetic studies.

The most informative application of PET to pharmacokinetic studies entails the simultaneous acquisition of two sets of data as follows. A series of tomographic images are acquired for a total period of 1 to 2 hours after bolus intravenous administration of a tracer dose of the radiolabeled drug, which typically ranges from 100 to 1,000 MBq, together with the usual dose of unlabeled drug. The time over which individual images are measured generally increases during the course of the experiment and can range from 5 seconds to 10 minutes, as dictated by the combined rates of radioactive decay of the tracer and elimination of the drug from the body. The time-averaged concentration of radiotracer in discrete regions of interest within the image are used to construct time-activity curves. Arterial blood is also serially collected from the patient throughout the experiment for independent measurement of the radiotracer concentration in whole blood or plasma by liquid scintillation counting. The empirical kinetic model that best relates the time course of radiotracer in tissue to its concentration in plasma is then identified.117

In addition to its inherent expense and technical complexities, PET has a number of other distinct limitations when applied to the study of the tissue pharmacokinetics of a drug. The time period over which a radiolabeled drug can be monitored after administration of a tracer dose is effectively limited to three to four times the half-life of radioactive decay. The maximum duration of an experiment is therefore only 90 minutes for a 11C-labeled drug and 6 to 8 hours when 18F is used as the radiotracer. For some drugs, this may not be enough time to adequately define the time course of the uptake or decline of radiotracer in the tumor. Furthermore, expecting patients to remain within the confines of the PET camera for more than 60 to 90 minutes is unreasonable, and demands on the instrument for routine diagnostic use may also represent a significant factor that limits the duration of an experiment.

Another important deficiency is that PET measures total radioactivity without distinguishing alterations in the chemical structure of the labeled molecule. Because PET cannot distinguish the parent drug from metabolites that retain the label, or free from protein-bound drug, it may not be suitable for studying the distribution of drugs that are extensively metabolized or highly protein bound. These factors need to be taken into consideration for each individual agent when analyzing and interpreting kinetic data from PET studies.117 Nevertheless, PET can provide highly informative insights and answer the critical question of whether radiotracer originating from the drug accumulates within a tumor.

Dynamic PET imaging has been used to study the distribution of the radiolabeled forms of several clinically approved and investigational anticancer agents to human brain tumors, including [11C]-BCNU,118, 119 [13N]-cisplatin,120 5-[18F]-fluorouracil,114 and [11C]-temozolomide.117 PET has been used to demonstrate that the initial uptake of radioactivity originating from [11C]-temozolomide in plasma was seven times greater in brain tumors than in normal brain tissue, with the kinetic behavior in these two regions becoming almost indistinguishable by 30 minutes after dosing.117 In consideration of these results, the markedly greater accumulation of radiolabeled agent in the tumor than in normal brain tissue, clearly evident in the PET images, was attributed to increased drug delivery due to a breakdown of the BBB within the tumor. The suggestion that the initial uptake of temozolomide from plasma could be an important determinant of its efficacy against human brain tumors has potential clinical significance.


Patients with primary or metastatic brain tumors are usually excluded from initial phase I clinical trials of investigational new anticancer drugs because of difficulties in differentiating potential indications of drug-related neurotoxicity from complications associated with the tumor. Furthermore, the additional supporting medications used in the clinical management of patients with brain tumors, especially antiseizure drugs and corticosteroids, could suppress the presentation of symptoms indicative of drug-related neurotoxicity. Until recently, the maximum tolerated dose of an investigational chemotherapeutic agent determined in adult patients with systemic solid tumors was used directly in phase II trials to assess clinical activity against brain tumors, without provisions to further refine the dose. It has become recognized that this practice often resulted in significantly undertreating patients with CNS malignancies as a consequence of pharmacokinetic interactions with concurrent medications that can enhance the elimination of anticancer drugs from systemic circulation as well as impede their access to the CNS.121, 122

Interactions Affecting the Elimination of Anticancer Agents

As indicated in Chapter 3, metabolism represents a quantitatively significant pathway of elimination for many anticancer chemotherapeutic agents. That certain classes of compounds can induce or suppress the expression of CYP450 enzymes and thereby alter the extent to which other drugs are metabolized by these pathways is well known.121, 122 In addition, competitive inhibition could result from the concurrent administration of two or more drugs that are substrates for the same CYP450 isozyme. These effects may not result in a clinically significant alteration in the pharmacokinetic behavior of most drugs. However, anticancer agents are typically administered at relatively high doses, close to the threshold of tolerability, and thus may utilize a greater capacity of the elimination pathways than drugs given for other indications. Accordingly, the potential for clinically significant pharmacokinetic interactions are greater for chemotherapeutic agents than for other classes of drugs.123

Antiseizure drugs such as phenytoin, carbamazepine, and phenobarbital, which are commonly administered on a long-term basis to brain tumor patients, are potent inducers of many of the most important CYP450 enzymes involved in drug metabolism, such as CYP2C8, CYP2C9, and CYP3A4.124 Patients who are being treated with these medications exhibit increased systemic clearance of epipodophyllotoxins,125, 126 vinca alkaloids,127 taxanes,128 and the camptothecins.129, 130 Administering standard doses of these chemotherapeutic agents together with an enzyme-inducing antiseizure drug results in lower plasma concentrations of the anticancer drug and reduced systemic toxicity. In consideration of the potential for pharmacokinetic interactions such as these, phase I studies are now being routinely designed to independently establish the maximum tolerated dose of new chemotherapeutic agents in brain tumor patients stratified according to whether or not they are receiving enzyme-inducing antiseizure drugs. Table 21.1 lists approved anticancer drugs for which pharmacokinetic interactions with enzyme-inducing antiseizure drugs have been evaluated in a clinical study in brain cancer patients and their effect on anticancer drug clearance.

Corticosteroids such as dexamethasone are widely used for the treatment of vasogenic brain edema and increased intracranial pressure in patients with brain tumors. Dexamethasone induces CYP3A4 by a pretranslational mechanism involving a glucocorticoid-responsive sequence in the promoter of the gene encoding the enzyme.131 In addition, dexamethasone is a potent inducer of CYP2C8 and CYP2C9. Preclinical studies have shown that the pharmacokinetic behavior of cyclophosphamide and ifosfamide were markedly affected by pretreatment with corticosteroids.132-134 Docetaxel metabolism was shown to be induced by dexamethasone in vitro, and decreased plasma concentrations of the drug have been demonstrated in a rodent model.135, 136 Corticosteroids appear to have little or no effect on the pharmacokinetics of the chloroethylnitrosourea alkylating agents.137 In addition to interactions originating from the induction of drug-metabolizing enzymes, the potential also exists for diminished elimination of an anticancer agent in brain cancer patients due to competitive inhibition of a major drug-metabolizing enzyme by a supporting medication.


Anticancer Agent

Infusion Time (h)

Dose (mg/m2)

Total Body Clearancea(l/hr/m2)

Difference (%)






















































aMean or median values.
bDose and clearance values are not normalized to body surface area.

Therapies That Modulate Drug Distribution to the Brain

Corticosteroids have a well-established effect on decreasing the permeability of the BBB and BTB to a wide variety of molecules. Preclinical studies have shown that dexamethasone significantly reduces the transport through the BBB of water,138 small organic molecules with molecular weights in the 100 to 350 range,139, 140 and macromolecules such as horseradish peroxidase.141 Treatment with corticosteroids diminishes the permeability of experimental brain tumors, brain tissue immediately adjacent to tumor, and normal brain tissue distant to tumor, but the effect is most pronounced within the tumor itself.142These findings have been corroborated in studies of brain tumor patients. A marked reduction in the permeability of tumor and normal brain tissue to 82Rb, as measured by PET imaging, was evident within 6 hours after the administration of dexamethasone by bolus injection to patients, and the effect persisted for at least 24 hours.143 The magnitude of the decreased uptake of 82Rb ranged from 6 to 48%.144 Similar results have been observed in other studies using CT scanning145 and MRI.146, 147 The effect of corticosteroids on the uptake of systemically administered chemotherapeutic agents has been evaluated in nude mice bearing intracranially implanted xenografts of human glioma. Steroid administration decreased the amount of carboplatin,148 cisplatin,149 and methotrexate150 in the tumor and surrounding brain tissue by 20 to 40%. The extent to which the distribution of an anticancer agent to brain tumors is affected by corticosteroids in humans, however, remains to be determined.

The effect of radiotherapy on the integrity of the BBB and consequent penetration of drug into brain tumors remains an open area of investigation. Conflicting observations have been reported and may be attributed to marked differences in radiation treatment protocols, the methods used to assess the impact of the treatments on the function of the BBB, and the time course of changes in vascular physiology.151 Consideration of the evidence derived from preclinical investigations indicates that the BBB in normal tissue becomes more permeable shortly after delivery of radiation according to the standard regimen of 5 days per week for 6 weeks.152 This is entirely consistent with the finding that P-glycoprotein labeling decreased by 60% in the endothelial cells of brain vessels in a rodent after irradiation.153 Slowly progressive alterations in the microvasculature of the irradiated tissue eventually result in decreased permeability.

The optimal schedule for delivering chemotherapy when used in combination with radiotherapy for treating brain tumors has not been conclusively established. The accumulation of methotrexate in a brain tumor model in mice was impaired by delivering radiation either before or concurrently with the systemically administered drug and resulted in shorter survival times; this suggests that chemotherapy should be given before radiation treatments.154 In contrast, clinical observations indicating that a 30- to 40-Gy dose of radiation increased the permeability of the BBB within the irradiated tumor by 74% but only by 24% in normal surrounding tissue, as assessed by scintigraphy, were the basis for advocating the administration of anticancer drugs after radiotherapy.155 Another study involving pediatric leukemia patients found no difference in the CSF-to-plasma concentration ratio of chemotherapeutic agents when given before, during, or after radiotherapy.156 Finally, concurrent administration of temozolomide and radiation in patients with newly diagnosed glioblastoma has been demonstrated to extend progression-free survival and overall survival, suggesting clinical benefit for this strategy.2 Because clinical trials for many new anticancer agents in brain tumor patients are now conducted in both the preradiation and postradiation periods, comparison of drug distribution into the CNS and brain tumors for these settings may be warranted.


Treatment strategies for the most common type of primary brain tumor, malignant glioma, are rapidly expanding. The persistence of normal BBB near the proliferating edge of the tumor, coupled with normalization of other areas of the BBB with treatment, emphasize the importance of strategies aimed at improving drug delivery across the BBB and to the tumor. In addition to the methods discussed in this review, cytotoxic drugs specifically designed for BBB penetration represent an important class of therapies to be assessed in the future. Methods for evaluating the success of these strategies are under development. Pharmacokinetic studies have assumed great importance in the development of antineoplastic therapy for hematologic and solid malignancies. Although application of the same principles to studies of brain tumor therapies is a relatively recent development and represents a unique set of challenges, these correlative studies add valuable information for the assessment of new brain tumor therapies. Moreover, with the emergence of cytostatic therapies for cancer, the traditional radiographic endpoints are insufficient, and these evaluative methods are likely to become surrogate endpoints in future clinical trials of these therapies. Finally, drug interactions have assumed great importance in brain tumor clinical trials owing to the recognition that many common supporting medications used in this patient population affect the metabolism of cytotoxic drugs through induction of the CYP450 enzyme family. The development of noninvasive methods that more readily facilitate evaluation of drug distribution and accumulation in local tissue and tumor is a fundamental challenge for the future. The availability of such techniques will allow efficient assessment of promising agents for the treatment of malignant gliomas.


1. CBTRUS. Statistical Report: Primary Brain Tumors in the United States, 1997–2001. Central Brain Tumor Registry of the United States, 2004.

2. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352:987.

3. Westphal M, Hilt DC, Bortey E, et al. A phase 3 trial of local chemotherapy with biodegradable carmustine (BCNU) wafers in patients with primary malignant glioma. Neurooncol 2003;5:79.

4. Stewart LA. Chemotherapy in adult high-grade glioma: a systematic review and meta-analysis of individual patient data from 12 randomized trials. Lancet 2002;359:1011.

5. Phillips PC. Antineoplastic drug resistance in brain tumors. Neurol Clin 1991;9:383.

6. Greig NH. Optimizing drug delivery to brain tumors. Cancer Treat Rev 1987;14:1.

7. Sipos EP, Brem H. New delivery systems for brain tumor therapy. Neurol Clin 1995;13:813.

8. Pardridge WM, Oldendorf WH, Cancilla P, et al. Blood-brain barrier: interface between internal medicine and the brain. Ann Intern Med 1986;105:82.

9. Zlokovic BV, Apuzzo ML. Strategies to circumvent vascular barriers of the central nervous system. Neurosurgery 1998;43:877.

10. Crone C. The blood-brain barrier: a modified tight epithelium. In: Succkling AJ, Rumsby MG, Bradbury MWB, eds. The Blood-Brain Barrier in Health and Disease. Chichester, England: Ellis Horwood, 1986:17.

11. Muldoon LL, Pagel MA, Kroll RA, et al. A physiological barrier distal to the anatomic blood-brain barrier in a model of transvascular delivery. Am J Neuroradiol 1999;20:217.

12. Fishman RA. Cerebrospinal Fluid in Diseases of the Nervous System. Philadelphia: WB Saunders, 1992:43.

13. Greig NH, Momma S, Sweeney DJ, et al. Facilitated transport of melphalan at the rat blood-brain barrier by the large neutral amino acid carrier system. Cancer Res 1987;47:1571.

14. Henson JW, Cordon-Cardo C, Posner JB. P-glycoprotein expression in brain tumors. J Neurooncol 1992;14:37.

15. Fenart L, Buee-Scherrer V, Descamps L, et al. Inhibition of P-glycoprotein: rapid assessment of its implication in blood-brain barrier integrity and drug transport to the brain by an in vitro model of the blood-brain barrier. Pharm Res 1998;15:993.

16. Tsuji A. P-glycoprotein–mediated efflux transport of anticancer drugs at the blood-brain barrier. Ther Drug Monit 1998; 20:588.

17. Neuwelt EA. Mechanism of disease: the blood-brain barrier. Neurosurgery 2004;54:131.

18. Vick NA, Khandekar JD, Bigner DD. Chemotherapy of brain tumors: the “blood brain barrier” is not a factor. Arch Neurol 1977;34:523.

19. Waggener JD, Beggs JL. Vasculature of neural neoplasms. Adv Neurol 1976;15:27.

20. Steinhoff H, Grumme T, Kazner E, et al. Axial transverse computerized tomography in 73 glioblastomas. Acta Neurochir 1978; 42:45.

21. Blasberg RG, Groothuis DR. Chemotherapy of brain tumors: physiological and pharmacokinetic considerations. Semin Oncol 1986;13:70.

22. Chamberlain MC, Murovic JA, Levin VA. Absence of contrast enhancement on CT brain scans of patients with supratentorial malignant gliomas. Neurology 1988;38:1371.

23. DeAngelis LM. Cerebral lymphoma presenting as a nonenhancing lesion on computed tomographic/magnetic resonance scan. Ann Neurol 1993;33:308.

24. Brooks DJ, Beaney RP, Thomas DG. The role of positron emission tomography in the study of cerebral tumors. Semin Oncol 1986;13:83.

25. Orr RJ, Brada M, Flower MA, et al. Measurements of blood-brain barrier permeability in patients undergoing radiotherapy and chemotherapy for primary cerebral lymphoma. Eur J Cancer 1991;27:1356.

26. Rapoport SI, Fredericks WR, Ohno K, et al. Quantitative aspects of reversible osmotic opening of the blood-brain barrier. Am J Physiol 1980;238:R421.

27. Jain RK. Transport of molecules in the tumor interstitium: a review. Cancer Res 1987;47:3039.

28. Yamada K, Ushio Y, Hayakawa T, et al. Distribution of radiolabeled 1-(4-amino-2-methyl-5-pyrimidinyl)methyl-3-(2-chloroethyl)-3-nitrosourea hydrochloride in rat brain tumor: intraarterial versus intravenous administration. Cancer Res 1987;47:2123.

29. Bullard DE, Bigner SH, Bigner DD. Comparison of intravenous versus intracarotid therapy with 1,3-bis(2-chloroethyl)-1-nitrosourea in a rat brain tumor model. Cancer Res 1985;45: 5240.

30. Fine HA, Dear KB, Loeffler JS, et al. Meta-analysis of radiation therapy with and without adjuvant chemotherapy for malignant gliomas in adults. Cancer 1993;71:2585.

31. Larner JM, Phillips CD, Dion JE, et al. A phase 1-2 trial of superselective carboplatin, low-dose infusional 5-fluorouracil and concurrent radiation for high-grade gliomas. Am J Clin Oncol 1995;18:1.

32. Stewart DJ, Grahovac Z, Hugenholtz H, et al. Combined intraarterial and systemic chemotherapy for intracerebral tumors. Neurosurgery 1987;21:207.

33. Shapiro WR, Green SB, Burger PC, et al. A randomized comparison of intra-arterial versus intravenous BCNU, with or without intravenous 5-fluorouracil, for newly diagnosed patients with malignant glioma. J Neurosurg 1992;76:772.

34. Kochi M, Kitamura I, Goto T, et al. Randomized comparison of intra-arterial versus intravenous infusion of ACNU for newly diagnosed patients with glioblastoma. J Neurooncol 2000;49:63.

35. Tamaki M, Ohno K, Niimi Y, et al. Parenchymal damage in the territory of the anterior choroidal artery following supraophthalmic intracarotid administration of CDDP for treatment of malignant gliomas. J Neurooncol 1997;35:65.

36. Blacklock JB, Wright DC, Dedrick RL, et al. Drug streaming during intra-arterial chemotherapy. J Neurosurg 1986;64:284.

37. Cloughesy TF, Gobin YP, Black KL, et al. Intra-arterial carboplatin chemotherapy for brain tumors: a dose escalation study based on cerebral blood flow. J Neurooncol 1997;35:121.

38. Saris SC, Blasberg RG, Carson RE, et al. Intravascular streaming during carotid artery infusions: demonstration in humans and reduction using diastole-phased pulsatile administration. J Neurosurg 1991;74:763.

39. Gobin PY, Cloughesy TF, Chow KL, et al. Intraarterial chemotherapy for brain tumors by using a spatial dose fractionation algorithm and pulsatile delivery. Radiology 2001;218:724.

40. Osztie E, Varallyay P, Doolittle ND, et al. Combined intraarterial carboplatin, intraarterial etoposide phosphate, and iv cytoxan chemotherapy for progressive optic-hypothalamic gliomas in young children. Am J Neuororadiol 2001;22:818.

41. Doolittle ND, Miner ME, Hall WA, et al. Safety and efficacy of multicenter study using intraarterial chemotherapy in conjunction with osmotic opening of blood-brain barrier for the treatment of patients with malignant brain tumors. Cancer 2000; 88:637.

42. Kroll RA, Neuwelt EA. Outwitting the blood-brain barrier for therapeutic purposes: osmotic opening and other means. Neurosurgery 1998;42:1083.

43. Gumerlock MK, Neuwelt EA. Chemotherapy of brain tumors: innovative approaches. In: Morantz RA, Walsh JW, eds. Brain Tumors. New York: Marcel Dekker, 1994:763.

44. Neuwelt EA, Howieson J, Frenkel EP, et al. Therapeutic efficacy of multiagent chemotherapy with drug delivery enhancement by blood-brain barrier modification in glioblastoma. Neurosurgery 1986;19:573.

45. Sheikov N, McDonnold N, Vykhodsteva N, et al. Cellular mechanisms of blood-brain barrier opening induced by ultrasound in presence of microbubbles. Ultrasound Med Biol 2004; 30:979.

46. Prados MD, Schold SC Jr, Fine HA, et al. A randomized, double-blind, placebo-controlled, phase 2 study of RMP-7 in combination with carboplatin administered intravenously for the treatment of recurrent malignant glioma. Neuro-oncol. 2003; 5:96.

47. Fernandez-Hidalgo OA, Vanaclocha V, Vieitez JM, et al. High-dose BCNU and autologous progenitor cell transplantation given with intra-arterial cisplatinum and simultaneous radiotherapy in the treatment of high-grade gliomas: benefit for selected patients. Bone Marrow Transplant 1996;18:143.

48. Perez Martinez A, Quintero CV, Gonzalez VM, et al. High-dose chemotherapy with autologous stem cell rescue in children with high-risk and recurrent brain tumors. Ann Pediatr 2004;61:8.

49. Modak S, Gardner S, Dunkel IJ, et al. Tiothepa-based high-dose chemotherapy with autologous stem-cell rescue in patients with recurrent or progressive CNS germ cell tumors. J Clin Oncol 2004;22:1934.

50. Chen B, Ahmed T, Mannancheril A, et al. Safety and efficacy of high-dose chemotherapy with autologous stem cell transplantation for patients with malignant astrocytomas. Cancer 2004; 100:2201.

51. Abrey L, Childs BH, Paleologos N, et al. High-dose chemotherapy with stem cell rescue as initial therapy for anaplastic ologodendroglioma. J Neuro-Onc 2003;65:127.

52. Abrey LE, Moskowitz CH, Mason WP, et al. Intensive methotrexate and cytarabine followed by high-dose chemotherapy with autologous stem-cell rescue in patients with newly diagnosed primary CNS lymphoma: an intent-to-treat analysis. J Clin Oncol 2003;21:4151

53. Bakhshi S, North RB. Implantable pumps for drug delivery to the brain. J Neurooncol 1995;26:133.

54. Pinkel D, Woo S. Prevention and treatment of meningeal leukemia in children. Blood 1994;84:355.

55. Chamberlain MC, Kormanik PA, Barba D. Complications associated with intraventricular chemotherapy in patients with leptomeningeal metastases. J Neurosurg 1997;87:694.

56. Chamberlain MC, Kormanik P, Howell SB, et al. Pharmacokinetics of intralumbar DTC-101 for the treatment of leptomeningeal metastases. Arch Neurol 1995;52:912.

57. Ettinger LJ, Chervinsky DS, Freeman AI, et al. Pharmacokinetics of methotrexate following intravenous and intraventricular administration in acute lymphocytic leukemia and non-Hodgkin's lymphoma. Cancer 1982;50:1676.

58. Bleyer WA, Drake JC, Chabner BA. Neurotoxicity and elevated cerebrospinal-fluid methotrexate concentration in meningeal leukemia. N Engl J Med 1973;289:770.

59. Shapiro WR, Young DF, Mehta BM. Methotrexate: distribution in cerebrospinal fluid after intravenous, ventricular and lumbar injections. N Engl J Med 1975;293:161.

60. Blasberg RG, Patlak C, Fenstermacher JD. Intrathecal chemotherapy: brain tissue profiles after ventriculocisternal perfusion. J Pharmacol Exp Ther 1975;195:73.

61. Blasberg RG. Methotrexate, cytosine arabinoside, and BCNU concentration in brain after ventriculocisternal perfusion. Cancer Treat Rep 1977;61:625.

62. Chamberlain MC. Leptomeningeal metastases: a review of evaluation and treatment. J Neurooncol 1998;37:271.

63. Walter KA, Tamargo RJ, Olivi A, et al. Intratumoral chemotherapy. Neurosurgery 1995;37:1128.

64. Mak M, Fung L, Strasser JF, et al. Distribution of drugs following controlled delivery to the brain interstitium. J Neurooncol 1995; 26:91.

65. Wu MP, Tamada JA, Brem H, et al. In vivo versus in vitro degradation of controlled release polymers for intracranial surgical therapy. J Biomed Mater Res 1994;28:387.

66. Leong KW, D'Amore PD, Marletta M, et al. Bioerodible polyanhydrides as drug-carrier matrices, II: biocompatibility and chemical reactivity. J Biomed Mater Res 1986;20:51.

67. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. The Polymer-Brain Tumor Treatment Group. Lancet 1995;345: 1008.

68. Brem H, Mahaley MS Jr, Vick NA, et al. Interstitial chemotherapy with drug polymer implants for the treatment of recurrent gliomas. J Neurosurg 1991;74:441.

69. Brem H, Ewend MG, Piantadosi S, et al. The safety of interstitial chemotherapy with BCNU-loaded polymer followed by radiation therapy in the treatment of newly diagnosed malignant gliomas: phase I trial. J Neurooncol 1995;26:111.

70. Valtonen S, Timonen U, Toivanen P, et al. Interstitial chemotherapy with carmustine-loaded polymers for high-grade gliomas: a randomized double-blind study. Neurosurgery 1997;41:44.

71. Olivi A, Barker F, Tatter S, et al. Toxicities and pharmacokinetics of interstitial BCNU administrated via wafers: results of a phase I study in patients with recurrent malignant glioma. Proc Am Soc Clin Oncol 1999;18:142a.

72. Limentani S, Asher A, Fraser R, et al. A phase I trial of surgery, Gliadel and carboplatin in combination radiation therapy for anaplastic astrocytoma (AA) or glioblastoma multiforme (GBM). Proc Am Soc Clin Oncol 1999;18:151a.

73. Mrugala MM, Kesari S, Ramakrishna N, et al. Therapy for recurrent malignant glioma in adults. Expert Rev Anticancer Ther 2004;4:759.

74. Kunwar S. Convection enhanced delivery of IL13-PE38QQR for treatment of recurrent malignant glioma: presentation of interim findings from ongoing phase 1 studies. Acta Neurochir 2003; 88(Suppl):105.

75. Groothius DR. The blood-brain and blood tumor barriers: a review of strategies for increasing drug delivery. Neuro-oncol 2000;2:45.

76. Groothuis DR, Benalcazar H, Allen CV, et al. Comparison of cytosine arabinoside delivery to rat brain by intravenous, intrathecal, intraventricular and intraparenchymal routes of administration. Brain Res 2000;856:281.

77. Laske DW, Youle RJ, Oldfield EH. Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med 1997;3:1362.

78. De Lange EC, Danhof M, de Boer AG, et al. Methodological considerations of intracerebral microdialysis in pharmacokinetic studies on drug transport across the blood-brain barrier. Brain Res Brain Res Rev 1997;25:27.

79. Cserr HF. Physiology of the choroid plexus. Physiol Rev 1971; 51:273.

80. Fishman RA. Cerebrospinal Fluid in Diseases of the Nervous System. Philadelphia: WB Saunders, 1992:23.

81. Zamboni WC, Gajjar AJ, Mandrell TD, et al. A four-hour topotecan infusion achieves cytotoxic exposure throughout the neuraxis in the nonhuman primate model: implications for treatment of children with metastatic medulloblastoma. Clin Cancer Res 1998;4:2537.

82. Blaney SM, Poplack DG, Godwin K, et al. Effect of body position on ventricular CSF methotrexate concentration following intralumbar administration. J Clin Oncol 1995;13:177.

83. Donelli MG, Zucchetti M, D'Incalci M. Do anticancer agents reach the tumor target in the human brain? Cancer Chemother Pharmacol 1992;30:251.

84. Stewart DJ. A critique of the role of the blood-brain barrier in the chemotherapy of human brain tumors. J Neurooncol 1994; 20:121.

85. Johansen MJ, Newman RA, Madden T. The use of microdialysis in pharmacokinetics and pharmacodynamics. Pharmacotherapy 1997;17:464.

86. Mary S, Muret P, Makki S, et al. A new technique for study of cutaneous biology, microdialysis. Ann Dermatol Venereol 1999; 126:66.

87. Groth L. Cutaneous microdialysis: methodology and validation. Acta Derm Venereol 1996;197(Suppl):1.

88. Ungerstedt U, Pycock C. Functional correlates of dopamine neurotransmission. Bull Schweiz Akad Med Wiss 1974;30:44.

89. Muller M, Mader RM, Steiner B, et al. 5-Fluorouracil kinetics in the interstitial tumor space: clinical response in breast cancer patients. Cancer Res 1997;57:2598.

90. Hamani C, Luer MS, Dujovny M. Microdialysis in the human brain: review of its applications. Neurol Res 1997;19:281.

91. Landolt H, Langemann H. Cerebral microdialysis as a diagnostic tool in acute brain injury. Eur J Anaesthesiol 1996;13:269.

92. Benveniste H, Hansen AJ, Ottosen NS. Determination of brain interstitial concentrations by microdialysis. J Neurochem 1989; 52:1741.

93. Major O, Shdanova T, Duffek L, et al. Continuous monitoring of blood-brain barrier opening to Cr51-EDTA by microdialysis following probe injury. Acta Neurochir Suppl (Wien) 1990;51:46.

94. Westergren I, Nystrom B, Hamberger A, et al. Intracerebral dialysis and the blood-brain barrier. J Neurochem 1995;64:229.

95. Hogan BL, Lunte SM, Stobaugh JF, et al. On-line coupling of in vivo microdialysis sampling with capillary electrophoresis. Anal Chem 1994;66:596.

96. Chen A, Lunte CE. Microdialysis sampling coupled on-line to fast microbore liquid chromatography. J Chromatogr A 1995; 691:29.

97. De Lange EC, Bouw MR, Mandema JW, et al. Application of intracerebral microdialysis to study regional distribution kinetics of drugs in rat brain. Br J Pharmacol 1995;116:2538.

98. Devineni D, Klein-Szanto A, Gallo JM. In vivo microdialysis to characterize drug transport in brain tumors: analysis of methotrexate uptake in rat glioma-2 (RG-2)-bearing rats. Cancer Chemother Pharmacol 1996;38:499.

99. El-Gizawy SA, Hedaya MA. Comparative brain tissue distribution of camptothecin and topotecan in the rat. Cancer Chemother Pharmacol 1999;43:364.

100. Zamboni WC, Houghton PJ, Hulstein JL, et al. Relationship between tumor extracellular fluid exposure to topotecan and tumor response in human neuroblastoma xenograft and cell lines. Cancer Chemother Pharmacol 1999;43:269.

101. Palsmeier RK, Lunte CE. Microdialysis sampling in tumor and muscle: study of the disposition of 3-amino-1,2,4-benzotriazine-1,4-di-N-oxide (SR 4233). Life Sci 1994;55:815.

102. Ekstrom O, Andersen A, Warren DJ, et al. Evaluation of methotrexate tissue exposure by in situ microdialysis in a rat model. Cancer Chemother Pharmacol 1994;34:297.

103. Blochl-Daum B, Muller M, Meisinger V, et al. Measurement of extracellular fluid carboplatin kinetics in melanoma metastases with microdialysis. Br J Cancer 1996;73:920.

104. Ronquist G, Hugosson R, Sjolander U, et al. Treatment of malignant glioma by a new therapeutic principle. Acta Neurochir 1992;114:8.

105. He Q, Bhujwalla ZM, Maxwell RJ, et al. Proton NMR observation of the antineoplastic agent iproplatin in vivo by selective multiple quantum coherence transfer (Sel-MQC). Magn Reson Med 1995;33:414.

106. Rodrigues LM, Maxwell RJ, McSheehy PM, et al. In vivo detection of ifosfamide by 31P-MRS in rat tumours: increased uptake and cytotoxicity induced by carbogen breathing in GH3 prolactinomas. Br J Cancer 1997;75:62.

107. Wolf W, Waluch V, Presant CA. Non-invasive 19F-NMRS of 5-fluorouracil in pharmacokinetics and pharmacodynamic studies. NMR Biomed 1998;11:380.

108. Findlay MP, Leach MO. In vivo monitoring of fluoropyrimidine metabolites: magnetic resonance spectroscopy in the evaluation of 5-fluorouracil. Anticancer Drugs 1994;5:260.

109. Presant CA, Wolf W, Waluch V, et al. Association of intratumoral pharmacokinetics of fluorouracil with clinical response. Lancet 1994;343:1184.

110. Maxwell RJ. New techniques in the pharmacokinetic analysis of cancer drugs, III: nuclear magnetic resonance. Cancer Surv 1993; 17:415.

111. Presant CA, Wolf W, Albright MJ, et al. Human tumor fluorouracil trapping: clinical correlations of in vivo 19F-nuclear magnetic resonance spectroscopy pharmacokinetics. J Clin Oncol 1990;8:1868.

112. Kristjansen PE, Quistorff B, Spang-Thomsen M, et al. Intratumoral pharmacokinetic analysis by 19F-magnetic resonance spectroscopy and cytostatic in vivo activity of gemcitabine (dFdC) in two small cell lung cancer xenografts. Ann Oncol 1993;4:157.

113. Wolf W, Waluch V, Presant CA, et al. Pharmacokinetic imaging of gemcitabine in human tumors using noninvasive 19F-MRS. Proc Am Assoc Cancer Res 1999;40:384.

114. Kissel J, Brix G, Bellemann ME, et al. Pharmacokinetic analysis of 5-[18F]fluorouracil tissue concentrations measured with positron emission tomography in patients with liver metastases from colorectal adenocarcinoma. Cancer Res 1997;57:3415.

115. Tilsley DW, Harte RJ, Jones T, et al. New techniques in the pharmacokinetic analysis of cancer drugs, IV: positron emission tomography. Cancer Surv 1993;17:425.

116. Rubin RH, Fischman AJ. Positron emission tomography in drug development. Q J Nucl Med 1997;41:171.

117. Meikle SR, Matthews JC, Brock CS, et al. Pharmacokinetic assessment of novel anti-cancer drugs using spectral analysis and positron emission tomography: a feasibility study. Cancer Chemother Pharmacol 1998;42:183.

118. Tyler JL, Yamamoto YL, Diksic M, et al. Pharmacokinetics of superselective intra-arterial and intravenous 11C-BCNU evaluated by PET. J Nucl Med 1986;27:775.

119. Diksic M, Sako K, Feindel W, et al. Pharmacokinetics of positron-labeled 1,3-bis(2-chloroethyl)nitrosourea in human brain tumors using positron emission tomography. Cancer Res 1984; 44:3120.

120. Ginos JZ, Cooper AJ, Dhawan V, et al. [13N]-cisplatin PET to assess pharmacokinetics of intra-arterial versus intravenous chemotherapy for malignant brain tumors. J Nucl Med 1987; 28:1844.

121. Van Meerten E, Verweij J, Schellens JH. Antineoplastic agents: drug interactions of clinical significance. Drug Saf 1995; 12:168.

122. McLeod HL. Clinically relevant drug-drug interactions in oncology. Br J Clin Pharmacol 1998;45:539.

123. Kivisto KT, Kroemer HK, Eichelbaum M. The role of human cytochrome P450 enzymes in the metabolism of anticancer agents: implications for drug interactions. Br J Clin Pharmacol 1995;40:523.

124. Tanaka E. Clinically significant pharmacokinetic drug interactions between antiepileptic drugs. J Clin Pharm Ther 1999;24:87.

125. Baker DK, Relling MV, Pui CH, et al. Increased teniposide clearance with concomitant anticonvulsant therapy. J Clin Oncol 1992;10:311.

126. Rodman JH, Murry DJ, Madden T, et al. Altered etoposide pharmacokinetics and time to engraftment in pediatric patients undergoing autologous bone marrow transplantation. J Clin Oncol 1994;12:2390.

127. Villikka K, Kivisto KT, Maenpaa H, et al. Cytochrome P450-inducing antiepileptics increase the clearance of vincristine in patients with brain tumors. Clin Pharmacol Ther 1999;66:589.

128. Chang SM, Kuhn JG, Rizzo J, et al. Phase I study of paclitaxel in patients with recurrent malignant glioma: a North American Brain Tumor Consortium report. J Clin Oncol 1998;16:2188.

129. Zamboni WC, Gajjar AJ, Heidman RL, et el. Phenytoin alters the disposition of topotecan and N-desmethyltopotecan in a patient with medulloblastoma. Clin Cancer Res 1998;4:783.

130. Gilbert MR, Supko JG, Batchelor T, et al. Phase I clinical trial and pharmacokinetic study of irinotecan in adult patients with recurrent malignant glioma. Clin Cancer Res, 2003;9:2940.

131. Liddle C, Goodwin BJ, George J, et al. Separate and interactive regulation of cytochrome P450 3A4 by triiodothyronine, dexamethasone, and growth hormone in cultured hepatocytes. J Clin Endocrinol Metab 1998;83:2411.

132. Chang TK, Yu L, Maurel P, et al. Enhanced cyclophosphamide and ifosfamide activation in primary human hepatocyte cultures: response to cytochrome P-450 inducers and autoinduction by oxazaphosphorines. Cancer Res 1997;57:1946.

133. Brain EG, Yu LJ, Gustafsson K, et al. Modulation of P450-dependent ifosfamide pharmacokinetics: a better understanding of drug activation in vivo. Br J Cancer 1998;77:1768.

134. Yu LJ, Drewes P, Gustafsson K, et al. In vivo modulation of alternative pathways of P-450-catalyzed cyclophosphamide metabolism: impact on pharmacokinetics and antitumor activity. J Pharmacol Exp Ther 1999;288:928.

135. Marre F, Sanderink GJ, de Sousa G, et al. Hepatic biotransformation of docetaxel (Taxotere) in vitro: involvement of the CYP3A subfamily in humans. Cancer Res 1996;56:1296.

136. Kamataki T, Yokoi T, Fujita K, et al. Preclinical approach for identifying drug interactions. Cancer Chemother Pharmacol 1998; 42:S50.

137. Levin VA, Stearns J, Byrd A, et al. The effect of phenobarbital pretreatment on the antitumor activity of 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU), 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) and 1-(2-chloroethyl)-3-(2,6-dioxo-3-piperidyl-1-nitrosourea (PCNU), and on the plasma pharmacokinetics and biotransformation of BCNU. J Pharmacol Exp Ther 1979;208:1.

138. Reid AC, Teasdale GM, McCulloch J. The effects of dexamethasone administration and withdrawal on water permeability across the blood-brain barrier. Ann Neurol 1983;13:28.

139. Ziylan YZ, LeFauconnier JM, Bernard G, et al. Effect of dexamethasone on transport of alpha-aminoisobutyric acid and sucrose across the blood-brain barrier. J Neurochem 1988;51: 1338.

140. Ziylan YZ, Lefauconnier JM, Bernard G, et al. Regional alterations in blood-to-brain transfer of alpha-aminoisobutyric acid and sucrose, after chronic administration and withdrawal of dexamethasone. J Neurochem 1989;52:684.

141. Hedley-Whyte ET, Hsu DW. Effect of dexamethasone on blood-brain barrier in the normal mouse. Ann Neurol 1986; 19:373.

142. Shapiro WR, Hiesiger EM, Cooney GA, et al. Temporal effects of dexamethasone on blood-to-brain and blood-to-tumor transport of 14C-alpha-aminoisobutyric acid in rat C6 glioma. J Neurooncol 1990;8:197.

143. Jarden JO, Dhawan V, Moeller JR, et al. The time course of steroid action on blood-to-brain and blood-to-tumor transport of +82 Rb: a positron emission tomographic study. Ann Neurol 1989;25:239.

144. Jarden JO, Dhawan V, Poltorak A, et al. Positron emission tomographic measurement of blood-to-brain and blood-to-tumor transport of +82Rb: the effect of dexamethasone and whole-brain radiation therapy. Ann Neurol 1985;18:636.

145. Yeung WT, Lee TY, Del Maestro RF, et al. Effect of steroids on iopamidol blood-brain transfer constant and plasma volume in brain tumors measured with X-ray computed tomography. J Neurooncol 1994;18:53.

146. Andersen C, Astrup J, Gyldensted C. Quantitation of peritumoural oedema and the effect of steroids using NMR-relaxation time imaging and blood-brain barrier analysis. Acta Neurochir Suppl (Wien) 1994;60:413.

147. Ostergaard L, Hochberg FH, Rabinov JD, et al. Early changes measured by magnetic resonance imaging in cerebral blood flow, blood volume, and blood-brain barrier permeability following dexamethasone treatment in patients with brain tumors. J Neurosurg 1999;90:300.

148. Matsukado K, Nakano S, Bartus RT, et al. Steroids decrease uptake of carboplatin in rat gliomas—uptake improved by intracarotid infusion of bradykinin analog, RMP-7. J Neurooncol 1997;34:131.

149. Straathof CS, van den Bent MJ, Ma J, et al. The effect of dexamethasone on the uptake of cisplatin in 9L glioma and the area of brain around tumor. J Neurooncol 1998;37:1.

150. Neuwelt EA, Barnett PA, Bigner DD, et al. Effects of adrenal cortical steroids and osmotic blood-brain barrier opening on methotrexate delivery to gliomas in the rodent: the factor of the blood-brain barrier. Proc Natl Acad Sci USA 1982;79:4420.

151. Trnovec T, Kallay Z, Bezek S. Effects of ionizing radiation on the blood brain barrier permeability to pharmacologically active substances. Int J Radiat Oncol Biol Phys 1990;19:1581.

152. d'Avella D, Cicciarello R, Angileri FF, et al. Radiation-induced blood-brain barrier changes: pathophysiological mechanisms and clinical implications. Acta Neurochir Suppl (Wien) 1998;71: 282.

153. Mima T, Toyonaga S, Mori K, et al. Early decrease of P-glycoprotein in the endothelium of the rat brain capillaries after moderate dose of irradiation. Neurol Res 1999;21:209.

154. Remsen LG, McCormick CI, Sexton G, et al. Decreased delivery and acute toxicity of cranial irradiation and chemotherapy given with osmotic blood-brain barrier disruption in a rodent model: the issue of sequence. Clin Cancer Res 1995;1:731.

155. Qin DX, Zheng R, Tang J, et al. Influence of radiation on the blood-brain barrier and optimum time of chemotherapy. Int J Radiat Oncol Biol Phys 1990;19:1507.

156. Riccardi R, Riccardi A, Lasorella A, et al. Cranial irradiation and permeability of blood-brain barrier to cytosine arabinoside in children with acute leukemia. Clin Cancer Res 1998;4:69.