Chapter 7 Chemotherapy for Brain Tumors

Factors Influencing Delivery of Chemotherapy to the Brain

Blood–Brain Barrier

Treatment of brain tumors with systemic chemotherapy poses challenges unique to brain tumors. Concentrations of chemotherapeutic agents within central nervous system (CNS) depends on multiple factors, including ability of the agents to cross the blood–brain barrier (BBB), the volume of distribution of the drug in the brain parenchyma and the extent to which the drug is actively transported out of the brain.¹ Therefore many promising compounds fail in CNS drug development due to limited access to the target sites in the brain. Foremost is the BBB, which impedes delivery of adequate concentrations of most chemotherapeutic agents to the tumor; others include the brain tumor barrier (BTB), blood–cerebrospinal fluid (CSF) barrier, and brain–CSF barrier.² Paul Ehrlich first described the BBB in 1985 when he noted that all body tissues except the brain were stained when certain vital dyes were injected intravenously into animals.³

The BBB critically controls the passage of drugs or other compounds from the blood to the CNS and protects the brain from the foreign and undesirable molecules. The major component of the BBB is a monolayer of brain capillary endothelial cells. The restriction of brain penetration arises from the presence of tight junctions between adjacent endothelial cells and interaction between astrocytes and endothelial cells.⁴ In contrast to other blood vessels in the body, the endothelial cells of brain capillaries lack intercellular fenestrations, and have high electrical resistance and low ionic permeability rendering them relatively impermeable to many water-soluble compounds.^2,⁴ The principal route to cross the BBB is via the lipid-mediated transport of small nonpolar molecules through the BBB by passive diffusion or less frequently by catalyzed transport.⁵ For a drug to successfully reach the brain parenchyma requires uptake across the luminal (blood-facing) membrane into the endothelial cells, transport across the transcellular membrane, and finally efflux across the abluminal side (brain parenchyma–facing membrane) into the interstitial fluid. The key to successful chemotherapy of brain tumors is adequate drug delivery to the tumor-infiltrated brain around the tumor and the individual tumor cells. To cross the BBB, chemotherapeutic drugs administered systemically must be less than approximately 200 daltons in size, lipid soluble, not bound to plasma proteins, and minimally ionized.^2,⁴ As a result, there is a positive correlation between lipophilicity of the drug and its ability to cross the BBB.

Role of Steroids

Steroids are important in the management of patients with brain tumors particularly in patients with bulky disease and those who have hydrocephalus. Dexamethasone has the best CNS penetration of all the steroids and is most commonly used in practice.⁶ Steroids decrease CSF production and cerebral blood flow and help to reduce vasogenic edema associated with the tumor. However, the steroid can impair delivery of the chemotherapeutic agents to the tumor.⁷ The assessment of response to treatment can also be affected by steroid use. Steroids can potentially decrease the degree of gadolinium enhancement that is a surrogate for the leakiness of the blood vessels and decrease the measurable volume of the tumor.⁸ Guidelines for determining response criteria to chemotherapy now require that the patient be on the same or a lower steroid dose as compared to baseline before determining an objective response.^9,¹⁰

Mechanisms of Drug Resistance and Strategies to Overcome Resistance

Efflux Transporters

However, uptake can be lower than predicted for drugs as they are subject to extrusion from the brain by active BBB efflux transporters. Drug transporters belong to two major superfamilies, ABC (adenosine triphosphate binding cassette) and SLC (solute carrier) transporters.^11,¹² ABC transporters are integral membrane proteins, many of which are located in the plasma membrane are primary active transporters, and they couple ATP hydrolysis to active efflux of their substrates against concentration gradients.¹³ The most extensively studied BBB transporter of the ABC family is P-glycoprotein (P-gp) initially discovered in 1976,¹⁴ but members of the multidrug resistance-associated proteins (MRP),¹⁵ family and breast cancer resistance protein (BCRP)¹⁶ have also been identified in brain endothelial cells and choroid plexus epithelial cells. Anticancer agents were among the first drugs identified to be substrates of BBB efflux transporters, that is, of Pgp as well as MRPs and BCRP.

DNA Repair Enzymes

Methyl guanine methyl transferase (MGMT) is an enzyme that removes chloroethylation or methylation damage from the O(6) position of guanine following alkylating chemotherapy, and hence is involved in DNA-repair.¹⁷ Clinical response to alkylating agents such as temozolomide (TMZ) in GBM patients has been correlated to the activity of the MGMT repair enzyme.¹⁸ The MGMT gene may be silenced by methylation of its promoter that prevents repair of DNA damage and increases the lethal effect of chemotherapy. O(6)-benzylguanine (O(6)-BG) is an AGT substrate that inhibits AGT by suicide inactivation, and based on these findings, clinical trials with agents such as O6-benzylguanine (O6BG) added to alkylating agents have been pursued that deplete MGMT.^19,²⁰ Unfortunately, this approach has been limited by systemic toxicities to date, as the combined toxicity of O6BG and TMZ has required significant dose reductions in TMZ, the presumed active agent for cancer cell death.²⁰

Poly (ADP-ribose) polymerase-1 (PARP-1) is an enzyme that catalyzes the transfer of β-nicotinamide adenine dinucleotide (NAD+) to poly(ADP-ribose).²¹ PARP-1 enzyme catalyzes the synthesis of polymers for DNA repair after injury, and PARP-1 influences both direct repair and base excision repair of DNA after injury from alkylating agents or ionizing radiation and is a key enzyme in the DNA repair pathways complementary to and downstream of MGMT.²¹ Hence, the PARP-1 enzyme inhibition is an attractive target for glioma therapy. PARP-1 inhibitors have also been shown to overcome resistance to TMZ in both mismatch repair-proficient and -deficient glioma cells in culture, and numerous PARP-1 inhibitors are in clinical trials in patients with high-grade gliomas.^22,²³

Strategies to Improve Drug Delivery to Treat Brain Tumors

Most cytotoxic agents do not cross BBB, and conventional methods of drug delivery often results in low levels of drug to the brain; therefore, innovative treatments and alternative delivery techniques are needed. These have included intra-arterial drug administration, high-dose chemotherapy, the use of drug embedded in a controlled-release, biodegradable matrix delivery system, disruption of the BBB by hyperosmolar solutions or biomolecules and convection enhanced delivery.

Intra-Arterial Chemotherapy

The goal of this approach is to deliver chemotherapy intra-arterially so that the tissue perfused by that artery is exposed to higher plasma concentrations of the drug during the first passage through the circulation. The principal advantage to this approach is to maximize the amount of drug crossing through the BBB and minimize systemic side effects. Theoretical modeling suggests that intra-arterial infusion can produce a 10-fold increase in peak drug concentrations as compared to intravenous infusion.²⁴ However, two phase-III trials failed to show a survival benefit for intra-arterial chemotherapy (IAC).^25,²⁶ A large trial of over 300 patients with newly diagnosed malignant glioma was conducted by the Brain Tumor Study Group (BTSG) trial to assess the efficacy of IAC chemotherapy in which nutrients were randomly assigned to IAC or intravenous (IV) BCNU with or without IV 5-fluorouracil (5-FU) and radiation therapy (RT). This study was closed early when an interim analysis showed shorter survival times in patients receiving IAC.²⁶ The side effects of the IAC in these two studies included catheter-related complications such as bleeding, infection, thrombosis, treatment-related neurotoxicity, leukoencephalopathy, cortical necrosis, and ipsilateral blindness.^25,²⁶

Intra–Cerebrospinal Fluid Chemotherapy

Intra-CSF chemotherapy involves administration of drugs either into the lateral ventricle, usually through a surgically implanted subcutaneous device, such as an Ommaya reservoir ²⁷ or instilling the drug into the lumbar subarachnoid space (i.e., intrathecal therapy). The benefit of this approach is that small doses of chemotherapeutic agents given intrathecally can produce high concentrations within the CSF with minimal systemic toxicity. However, abnormal CSF flow and obstruction due to tumor or scarring from prior surgical interventions impair its utility in the treatment for primary brain tumors. Intrathecal administration of drugs has limited penetration into the brain parenchyma, and is generally employed in treatment or prophylaxis of leptomeningeal disease.²⁸ Side effects include increased risks of neurotoxicity (especially with radiation) and chemical meningitis.

Manipulating BBB Permeability or Methods to Cause BBB Disruption

Various agents have been used to modify BBB and/or BTB in an attempt to increase the drug concentration in the tumor.²⁹ Drug delivery to brain tumors can potentially be improved by increasing the permeability of the BBB with hyperosmolar solutions such as mannitol and vasoactive compounds such as bradykinin analogues that induce an osmotic opening of the BBB and BTB.^29,³⁰ Hyperosmolar solutions increase capillary permeability by temporarily opening the intercellular tight junctions of the brain endothelium that results in increased movement of water soluble substances. Complications with this approach include increased risk of stroke and seizures,³¹ and no clinical benefit has been demonstrated with this approach so far.³²

High-Dose Chemotherapy

High-dose chemotherapy (HDCT) is theoretically promising as increases the peak concentration of unbound drug in the circulation and can result in greater transfer of drug across the BBB. The associated myelosuppression seen with this approach requires use of autologous hematopoietic cell rescue and treatment-related morbidity is substantial. The survival using this approach is similar to that achieved by conventional chemotherapy or targeted therapy and its use remains investigational.^33,³⁴

Wafers/Implantable Polymers

Surgical implantation of solid-phase reagents permits constant drug delivery into the tumor without significant systemic or local side effects. The most commonly used “wafer” is a copolymer matrix with carmustine that is implanted directly into the tumor resection cavity at the time of surgical intervention and has been approved for use in patients with malignant gliomas. This therapy has been approved for patients with newly diagnosed and recurrent high-grade gliomas.^35,³⁶

Convection-Enhanced Delivery

Convection-enhanced delivery (CED) involves direct intratumoral infusion with various chemotherapeutic drugs and has been designed to use pharmacological agents that would not normally cross the BBB, and this approach is particularly useful for the delivery of large molecules.³⁷ Drugs are delivered through one to several catheters placed stereotactically directly within the tumor mass or around the tumor or the resection cavity and it allows distribution of substances throughout the interstitium via positive-pressure infusion.³⁸ Several classes of drugs are amenable to this technology including chemotherapeutics or targeted drugs.³⁹ Two multicenter randomized controlled trials in patients with recurrent GBM (PRECISE and TransMID) demonstrated that CED of agents was safe and well tolerated.⁴⁰ However, no survival benefit was seen in PRECISE, a phase III trial to assess the efficacy of this approach in patients with GBM upon first relapse compared to treatment with carmustine wafers.⁴⁰ Results of the TransMID trial have not been presented or published yet.

Various CNS Tumors

Low-Grade Gliomas

Low-grade gliomas (LGGs) comprise approximately 20% of CNS glial tumors with approximately 1800 new cases diagnosed each year in the United States.⁴¹ Oligodendrogliomas represent 3.7% of all primary brain and CNS tumors.⁴¹ Patients with LGGs typically present between the second and fourth decades of life. The optimal role of surgical resection in the long-term outcome of patients with LGG remains controversial, and the debate about the effect on outcome of its timing and extent persists. Nevertheless, surgery continues to be indispensable to provide tissue for histopathologic diagnosis and importantly molecular characterization that is prognostic and helps determine therapy approach. Retrospective studies have shown that more extensive resection rather than simple debulking is more beneficial and that greater than 99% resection yields improved overall survival (OS) and progression-free survival (PFS).⁴²

Radiation Therapy

The value of RT in managing LGGs is controversial. This is due to prolonged natural history of LGGs and these patients are more likely to live long enough to suffer from the late effects of RT. In addition, dose of RT to treat LGGs is not clear. The most commonly used RT for the treatment of LGGs is 54 Gy with a range of 45 to 60 Gy, based on results of the European Organization for Research and Treatment of Cancer (EORTC) trial 22844 ⁴³ and the North Central Cancer Treatment Group/Radiation Therapy Oncology Group(RTOG)/Eastern Cooperative Oncology Group study.⁴⁴

In the EORTC trial, there was no significant difference in OS and PFS in patients of LGG treated with 59.4 Gy in 33 treatments or 45 Gy in 25 treatments.⁴³ In the multigroup trial there was no survival benefit of using 64.8 Gy compared to 50.4 Gy.⁴⁴ A higher dose of RT (64.8 Gy) resulted in higher rates of radiation necrosis, and consequently doses above 60 Gy are avoided in this patient group.⁴⁴

Moreover, the benefit of RT is limited to improvement in PFS without translating into any improvement in OS as was demonstrated by the EORTC trial 22845.⁴⁵ This study evaluated the role of up-front RT versus observation in LGG. A total of 311 patients were treated with immediate RT (54 Gy in 6 weeks) or no therapy until progression.⁴⁵ Up-front RT significantly prolonged the median PFS (5.4 years vs. 3.7 years), but did not result in improvement in OS (7.4 years vs. 7.2 years).⁴⁵ This suggests that radiation may have a comparable effect whether it is administered early or at subsequent tumor progression.

Chemotherapy

There is no level 1 evidence that postoperative chemotherapy significantly prolongs survival in patients with LGGs. The RTOG 98-02 was a three-arm trial in which 111 patients with a favorable prognosis (age <40 years and gross tumor resection) were followed with observation following surgery. A total of 251 patients with an unfavorable prognosis (age ≥40 years or those who had subtotal resection or biopsy only) were randomized to receive RT (54 Gy in 30 fractions) followed by six cycles of procarbazine, lomustine, and vincristine (PCV), or the same dose of RT only. At the last update of the trial presented at the American Society of Clinical Oncology meeting in 2008, PFS was increased with adjuvant PCV chemotherapy; however, the OS was similar in the two groups.⁴⁶ However, after 2 years, the addition of chemotherapy to RT did result in a significant OS and PFS advantage suggesting a delayed advantage to chemotherapy.⁴⁶

Recently, TMZ has been increasingly been used to treat this patient population. In a retrospective review of 149 patients with LGGs treated with TMZ, a partial response (PR) rate of 15% and a minor response (MR) rate of 38% were reported.⁴⁷ In addition stable disease (SD) was seen in 37% and PD in 10% with a median PFS of 28 months. Tissue from 86 patients showed that codeletion of 1p/19q was associated with a significantly higher response rate (RR), a longer response to TMZ, especially with 1p/19q codeletion improved PFS and OS.⁴⁷ These results provide strong evidence that LGGs respond to TMZ. Besides LOH 1p/19q, methylation status of the MGMT promoter (MGMTP) predicts response to TMZ in LGG.⁴⁸ The LGG patients with methylation of MGMTP had an improved PFS compared to those with unmethylated MGMTP when treated with TMZ (p < 0.0001).⁴⁸

Malignant Glioma

Malignant gliomas (MG) or high-grade gliomas (HGG) include WHO grade IV gliomas, also known as GBM, and WHO grade III gliomas referred to as anaplastic gliomas (AG) (anaplastic astrocytoma [AA], anaplastic oligodendroglioma [AO], and anaplastic oligoastrocytoma [AOA]).⁴⁹ The goals of surgery are to establish a histological diagnosis and relieve mass effect. Biopsy alone is used in situations where the lesion is not amenable to resection, or the patient’s overall clinical condition will not permit surgery. However, maximal surgical resection while preserving neurological function is preferred.

Glioblastoma

Radiation

Even patients who undergo a gross total resection of their glioblastoma (GBM) have a high recurrence rate, and for over three decades adjuvant radiation therapy has been the standard approach for GBMs. The efficacy of radiation was initially established in the 1970s in a trial of over 300 patients with an AG addition of adjuvant whole-brain radiation therapy (WBRT) to surgical resection resulted in increased median survival from 14 to 36 weeks.⁵⁰ A seminal analysis of patients treated in the previous Brain Tumor Cooperative Group Trials had established the standard radiation dose to be 60 Gy in the late 1970s,⁵¹ and dose escalation above 60 Gy with WBRT has not been shown to provide further additional benefit.^52,⁵³ Predominant mode of recurrence in patients with high-grade gliomas treated with radiation has been local failure within 2 cm of the enhancing tumor.⁵⁴^–⁵⁶ Serious side effects of WBRT such as progressive and irreversible radiation necrosis, with accompanying small blood vessel injury, and demyelination led to the utilization of involved field radiation therapy (IFRT) as the standard approach for adjuvant RT to minimize toxicity. The pattern of treatment failure seen with IFRT are similar to those seen with WBRT and are mostly local failures within 2 cm of the initial tumor.⁵⁵ In the United States, the RTOG treatment volumes generally used that deliver a 46-Gy dose to the peritumoral edema with a 2-cm margin and a 14-Gy boost to the enhancing tumor with a 2.5-cm margin. In Europe, a full 60 Gy are delivered to a 2- to 3-cm margin around the enhancing tumor without any field reduction. At present, T2 or fluid-attenuated, inversion-recovery magnetic resonance imaging (MRI) sequences are used to identify peritumoral edema, and the T1 sequence with contrast images is used to identify the enhancing portion of the tumor. If the tumor margin is defined upon contrast enhancement, typically a margin of 2.0 to 3.0 cm is used, and a margin of 1.0 to 2.0 cm is used if the RT field is defined by the T2-weighted MRI abnormality. Use of metabolic imaging such as positron emission tomography (PET), MR diffusion and MR perfusion, and MR spectroscopy are promising as they represent areas of activity that may need different treatment planning as compared to that defined by the traditional MRI sequences.^57,⁵⁸ However, they are still largely investigational at present, and are employed to define boost volumes rather than primary target volumes. Further advances in imaging will likely change the method of tumor delineation.

Intensity-modulated RT (IMRT) is a technique that utilizes software and modification of standard linear accelerator output to deliver varied intensity of radiation across each treatment field. IMRT is beneficial especially when the tumor is juxtaposed to radiation-sensitive structures. IMRT is increasingly used these days as it may reduce radiation-related adverse effects ⁵⁹ and can escalate the radiation dose delivered to the tumor. However, as of this writing, no proven benefit has been demonstrated by delivering doses in excess of 60 Gy.^59,⁶⁰

Chemotherapy

The benefit of adjuvant temozolomide (Stupp regimen) was established in a seminal phase III trial, when 573 patients with newly diagnosed GBM were randomly assigned to postoperative involved-field RT (60 Gy in daily 30 fractions) versus the RT plus concurrent temozolomide (TMZ) (75 mg/m ² daily up to 49 days) followed by up to six cycles of adjuvant TMZ (150 to 200 mg/m² daily for 5 days, every 28 days).⁶¹ This study demonstrated a benefit with adjuvant TMZ with a 2.5-month median improvement of OS (12.1 months for RT alone vs. 14.6 months for RT plus TMZ). At 2 years, 26.5% of patients treated with TMZ plus RT were alive, compared with 10.4% of patients treated with RT alone. This benefit was even more impressive at 5 years when 9.8% of patients treated with TMZ plus RT were alive compared to 1.9% patients treated with RT alone.⁶² However, this study did not include patients older than 70 years of age, which constitutes 20% of all patients with newly diagnosed GBMs (discussed below). It also excluded patients with low performance status who were not independent in activities of daily living, a group constituting at least 10% of all newly diagnosed patients with GBMs in which the treatment plan needs to be tailor made according to the patient’s ability to tolerate RT and or chemotherapy.

As previously mentioned, MGMT is an enzyme responsible for DNA repair following alkylating chemotherapy. During the course of tumor development, the MGMT gene may be silenced by methylation of its promoter, which prevents repair of DNA damage and increases the effectiveness of alkylating agents such as TMZ. MGMT was determined retrospectively from the tumor tissue of 206 patients and appeared to be a prognostic factor for improved survival and was predictive of benefit from chemotherapy.¹⁸ For those with MGMT methylation, the 2-year survival rates were 49% and 24% with combination therapy and with RT alone, respectively, while for those without MGMT methylation, the 2-year survival rates were 15% and 2%, respectively. Biodegradable wafers impregnated with carmustine (Gliadel® wafer) that function as a slow-release carrier system for local drug delivery implanted at the time of resection are approved therapy for patients with newly diagnosed MG.^35,³⁶ In a phase III trial, 240 newly diagnosed malignant glioma patients were randomized to placement of up to eight carmustine wafers or a placebo, followed by standard RT.^35,³⁶ Patients receiving carmustine polymer had a statistically significant increase in median survival (13.9 vs. 11.6 months). This difference in survival, however, was not statistically significant when the analysis was restricted to GBM.^35,³⁶ Additional toxicities with Gliadel included increase in the incidence of CSF leak and intracranial hypertension compared to placebo.^35,³⁶

Bevacizumab is a monoclonal antibody that binds vascular endothelial growth factor (VEGF), which plays a critical role in the development of the abnormal vasculature observed in tumors including malignant gliomas.⁶³ Two phase II trials that have been reported evaluating the addition of Bevacizumab to standard RT and TMZ (Stupp regimen).^64,⁶⁵ In a study conducted at Duke, a total of 125 patients with newly diagnosed GBM received standard radiation therapy, TMZ, and bevacizumab. At a 21-month follow-up, the median PFS of 13.8 months and median OS of 21.3 months was reported.⁶⁵ These results are similar to the median PFS and OS of 13.6 and 19.6 months, respectively, in a phase II trial of bevacizumab plus TMZ and RT in patients with newly diagnosed GBM reported by Albert Lai and colleagues.⁶⁴ These compare with PFS and OS of 7.6 and 21.1 months in the University of California, Los Angeles/KPLA control cohort.⁶⁴ These two studies demonstrate that patients treated with BEV and TMZ during and after RT may show improved PFS without improved OS compared to the historical control group. The two phase III ongoing studies, RTOG 0825 (NCT00884741) and Hoffmann-La Roche Study (NCT00943826) will help answer the question of whether adding bevacizumab to TMZ and radiotherapy will improve survival of patients with GBM.

Recurrent Glioblastoma

Progression versus Pseudoprogression

Despite recent advances in therapeutics, most patients with GBM develop tumor recurrence after the above therapy. Recurrence is suspected when a previously stable patient develops new neurologic signs and symptoms or when surveillance MRI with gadolinium imaging shows increased tumor size or new enhancement usually accompanied with increased edema. However, clinical and imaging changes may be due to complications such as infection, a decrease in steroid use or radiation necrosis (also referred to as “pseudoprogression”). Radiation necrosis is a well-known late effect of RT of the brain that can mimic tumor recurrence. Pseudoprogression is a similar effect of transient increase in tumor enhancement that has been described after combined chemoradiotherapy and that occurs more rapidly and dramatically than after radiation alone. Pseudoprogression has been noted to occur between 20% and 30% of the cases in recent series.⁶⁶^–⁶⁸ It has been suggested that “pseudoprogression may occur more frequently in patients with methylation of the MGMT promoter as it increases the effect of chemoradiotherapy on residual tumor and that this translates to the transient worsening of imaging characteristics.⁶⁹ In the same series, survival in patients with pseudoprogression was significantly better than survival in those whose scans were initially stable (38 months vs. 20.2 months).⁶⁹ Various novel imaging modalities such as magnetic resonance perfusion with or without spectroscopy, and PET are used to help distinguish between pseudoprogression and true early progression of disease, but are not always reliable, and none has yet been widely accepted as standard practice. In most cases, repeat imaging is helpful to distinguish between the two, while in select cases surgery may be necessary to relieve mass effect and obtain a tissue diagnosis.

Therapeutic Options

When a tumor reaches a certain size, the requirements for oxygen and nutrients lead to the growth of new blood vessels and tumors can promote the formation of new vessels through the process of angiogenesis.⁷⁰ GBMs are among the most vascular tumors known and hence therapy directed against tumor-associated vasculature is a promising strategy.⁷⁰ Bevacizumab is a monoclonal antibody directed toward VEGF, and is the prototype of antiangiogenic agents in clinical use for treatment of GBM. Bevacizumab was approved by the Food and Drug Administration (FDA) for recurrent GB in the United States based on two trials of bevacizumab as a single agent or combined with irinotecan in recurrent GBM patients after initial treatment with chemoradiation and adjuvant TMZ. In a randomized, phase II clinical trial, 167 patients with recurrent GBM were treated with bevacizumab alone or bevacizumab in combination with irinotecan; there was no statistically significant difference in the median OS in the group treated with bevacizumab alone (9.2 months) compared to the those treated with combination of bevacizumab and irinotecan (8.7 months).⁷¹ The objective response rate was 25.9% in patients who received bevacizumab monotherapy, and there were no complete responses per the outside review. Median duration of response was 4.2 months among the responders and the 6-month PFS (PFS-6) was 36%. The second study by the National Cancer Institute involved 48 recurrent high-grade glioma heavily pretreated patients treated with bevacizumab alone.¹⁰ The objective response as determined by independent review was 19.6% and median duration of response was 3.9 months in responders.⁷² The FDA approved bevacizumab as a single agent based on improvement in objective response rate in these studies, although no increased survival was seen.

Potent anti-VEGF activity of bevacizumab results in normalization of permeable tumor vessels producing rapid and marked reduction in edema and contrast enhancement on neuroimaging.⁷³ This effect of rapid and dramatic improvements in MRI can occur within days of initiation of treatment with antiangiogenic agents such bevacizumab, cediranib, sunitinib, sorafenib, and aflibercept, and is partly a result of reduced vascular permeability to contrast agents rather than a true antitumor effect. These imaging changes can make evaluation of tumor response and progression difficult if one relies on commonly used MacDonald criteria of two-dimensional measurement of enhancing disease. In addition, a subset of patients treated with bevacizumab develop tumor recurrence observed as an increase in the nonenhancing component on T2-weighted/fluid-attenuated inversion recovery (FLAIR) sequences. This likely reflects a phenotypically invasive tumor recurrence pattern due to co-option of normal cerebral vessels and diffuse perivascular spread of tumor cells. Hence, the Response Assessment in Neuro-Oncology Working Group proposed a new standardized response criterion that takes into consideration the challenges of nonenhancing signal abnormality changes, pseudoprogression, and pseudoresponse.¹⁰

Carmustine polymer wafers (Gliadel) may prolong survival and has been approved for use after surgery in locally recurrent high-grade glioma.⁷⁴ A prospective, randomized phase-III trial demonstrated a modest increase in OS from 23 weeks in those patients who received placebo wafers compared to 31 weeks receiving Gliadel.⁷⁴ However, the study included recurrent low- and higher-grade gliomas, and the benefit in the GBM subgroup was smaller than in the whole cohort. This study predated the use of chemoradiation and adjuvant TMZ, and the benefit of this approach in recurrent GBM patients treated with prior TMZ is unclear.

Other options for patients who have contraindication to bevacizumab or prior to therapy with bevacizumab include rechallenge with alternative dosing schedules of TMZ.⁷⁵ One of the mechanisms of resistance to TMZ occurs through direct repair of DNA damage by MGMT enzyme, and an effective strategy to overcome such form of resistance is to deplete tumor cell MGMT. TMZ rechallenge with alternative doses and dosing schedules that deliver higher cumulative doses over prolonged periods of time can result in depletion of MGMT,⁷⁶ and has been shown to be directly toxic to endothelial cells.⁷⁷ Commonly used TMZ dosing schedules are 75 to 100 mg/m ² (21 days on/7 days off), 150 mg/m² (7 days on/7 days off), and 50 mg/m² daily dosing. Similar responses seen in patients with high and low levels of tumor MGMT support the rationale that these regimens may overcome MGMT-mediated resistance.⁷⁸

Other chemotherapy options for recurrent GBM include nitrosoureas (e.g., carmustine, fotemustine) either as single agent or in combination (most commonly used regimen, PCV) that have shown activity in previously treated patients.⁷⁹^–⁸¹ Other chemotherapeutic agents used in this patient population includes carboplatin, etoposide, and irinotecan, which have demonstrated modest efficacy as single agents or in combination regimens.⁸²^–⁸⁵ Recently in a randomized phase III trial of 325 recurrent GBM patients, lomustine was found to be superior to the pan-VEFG receptor inhibitor, cediranib.⁸⁶

Molecularly Targeted Therapy

In the past decade there has been substantial growth in the number of novel therapies due to increased understanding of the molecular pathways involved in glioma formation and progression. Malignant transformation in gliomas is often the result of the sequential accumulation of genetic aberrations and proliferation of growth factor signaling pathways that include the vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and platelet-derived growth factor (PDGF).⁸⁷

A number of agents that target VEGF have been developed including bevacizumab. Tyrosine kinase inhibitors that target VEGF pathway include cediranib,^88,⁸⁹ and adnectin.⁹⁰^–⁹² Small-molecule EGFR inhibitors such as gefitinib and erlotinib are well tolerated in patients with recurrent HGG, but responses have been disappointing.^93,⁹⁴ There are a number of agents that target different signal transduction pathways including PI3K/AKT/mTOR,^95,⁹⁶ RAF-MEK-ERK,^97,⁹⁸ PDGF,^99,¹⁰⁰, SRC,¹⁰¹ and PKC¹⁰² pathways are undergoing trials in patients with high-grade gliomas (Table 7-1). Results from clinical trials with most of these molecular-targeted therapies with the exception of bevacizumab have been disappointing so far. This is likely due to the complexity of the molecular abnormalities in recurrent high-grade gliomas, the redundancy of the signaling pathways, and the inability of many of these agents to cross the BBB.

TABLE 7-1 Selected Molecularly Targeted Agents in Clinical Trials in High-Grade Gliomas

Drug Name	Type of Drug	Targets
ABT-888	Tyrosine kinase inhibitor	PARP-1, PARP-2
Aflibercept	Soluble decoy receptor	VEGF-A,B, PlGF
AMG 102	Thrombospondin-1 mimetic peptide	FGFR, VEGFR2
Bevacizumab	Monoclonal antibody	VEGF-A
Brivanib	Monoclonal antibody	FGF pathway
Cediranib	Tyrosine kinase inhibitor	VEGFR1–3, PDGFRβ, c-Kit
Cetuximab (Erbitux)	Monoclonal antibody	EGFR
CT-322	Fibronectin (adnectin)-based inhibitor	VEGFR1–3
Dasatinib	Immunomodulatory and anti-inflammatory	PDGFRβ, BCR-ABL, c-Kit
Erlotinib (OSI-774)	Tyrosine kinase inhibitor	EGFR
Everolimus (RAD-001)	Tyrosine kinase inhibitor	mTOR
Gefitinib (ZD1839)	Tyrosine kinase inhibitor	EGFR
Imatinib	Tyrosine kinase inhibitor	PDGFRβ, Flt3, c-Kit
IMC-1121B	Monoclonal antibody	VEGR
Lapatinib (GW-572016)	Tyrosine kinase inhibitor	EGFR
Lenalidomide	Tyrosine kinase inhibitor	PDGFRβ, Src, BCR-ABL, c-Kit, EphA2
Lonafarnib (SCH-66336)	Farnesyl tranferase inhibitor	Ras
Pazopanib (GW786034)	Tyrosine kinase inhibitor	VEGFR1–3, PDGFRβ, c-Kit
Sorafenib	Tyrosine kinase inhibitor	VEGFR2,3, BRAF, PDGFRβ, c-Kit, Ras, p38α
Sunitinib	Tyrosine kinase inhibitor	VEGFR2, PDGFRβ, Flt3, c-Kit
Tandutinib (MLN518)	Tyrosine kinase inhibitor	COX-2
Temsirolimus (CC!-779)	Tyrosine kinase inhibitor	mTOR
Tipifarnib (R115777)	Farnesyl tranferase inhibitor	Ras
Vandetanib (ZD6474)	Tyrosine kinase inhibitor	VEGFR2, EGFR, RET
Vatalanib (PTK787)	Tyrosine kinase inhibitor	VEGFR1–3, PDGFRβ, c-Kit
XL-184	Tyrosine kinase inhibitor	VEGFR2, Met, RET, c-Kit, Flt3, Tie-2
XL-765	Tyrosine kinase inhibitor	mTOR, P13K