Introduction
High-grade or malignant astrocytomas consist of anaplastic astrocytoma (AA) and glioblastoma multiforme (GBM), and represent the most commonly diagnosed brain tumors in adults [1ā3]. These tumors account for 30%ā35% of all newly diagnosed brain tumors in patients 18 years of age or older and affect an estimated 10,000 to 12,000 new patients each year in the United States. The overall incidence rates for AA and GBM are 0.13 and 3.24 per 100,000 person-years, respectively. The age of onset is different between the two forms of high-grade astrocytoma, with AA arising in patients between the ages of 45 and 55 years, and GBM typically affecting an older cohort between 55 and 65 years of age. There is usually a slight preponderance of male patients, with a male to female ratio of approximately 1.8:1 in most series [2,3]. The prognosis for patients with AA and GBM remains poor, with overall median survivals of 30ā38 months and 14ā18 months, respectively, for patients that have received standard treatment.
AAs and GBM are classified by the World Health Organization (WHO) as grade III and grade IV astrocytomas, respectively (see Chapter 1) [4]. Pathologically, AAs are characterized by the presence of high cellularity, prominent cellular and nuclear pleomorphism, nuclear atypia, and mitotic activity. The histological features of GBM are similar to AA, but with more pronounced cellular and nuclear anaplasia and the presence of microvascular proliferation and/or necrosis. Both AA and GBM are highly infiltrative tumors and commonly arise in the cerebral hemispheres or in the deep white matter.
Overview of Initial Treatment
The most common form of initial treatment for a high-grade astrocytoma is surgical intervention. Indications for surgery include reducing tumor burden, alleviating mass effect, confirmation of the histological diagnosis, diversionary shunting procedures in selected cases, and the introduction of local antineoplastic agents [5ā7]. Recent advances in neurosurgical technology offer new approaches to tumor removal, such as frame-based and frameless stereotactic biopsy, intraoperative cortical mapping, neuronavigation, and the use of intraoperative magnetic resonance imaging (MRI) [6ā9]. These techniques allow the surgeon to delineate tumor margins more carefully and preserve the surrounding regions of eloquent brain (e.g., Broca's area, primary motor cortex) and delicate vascular structures, while performing a more aggressive and thorough tumor resection. Although it remains debated in the literature, most neurosurgeons recommend a near-total or gross-total resection, whenever possible, of all enhancing tumor volume and regionally infiltrated brain as defined on T2 MRI images. Gross-total tumor resection has been associated with longer overall and progression-free survival in several studies [10,11]. For tumors that are diffusely infiltrative or multifocal, a stereotactic biopsy is more likely to preserve neurological function than an attempt at resection and, in most cases, will be able to provide a histological diagnosis to guide further treatment.
Fractionated external-beam radiation therapy is an appropriate form of treatment for virtually all patients with AA and GBM, and is similar for both tumor types [1,5,12ā14]. Numerous randomized controlled trials have demonstrated a survival benefit for patients receiving surgical resection and irradiation in comparison to resection alone (approximately 34ā38 weeks vs 14ā18 weeks). The standard approach is administered in the early postoperative phase and uses initial radiation ports that encompass the T2-weighted target with a margin of 1ā3 cm, using a dose of approximately 4500ā4700 cGy in 180ā200 cGy daily fractions. After this portion has been completed, a ācone downā is performed, targeting the T1-weighted contrast-enhancing volume of the tumor with a 1ā3 cm margin, bringing the total dose to approximately 6000 cGy. Irradiation is performed over the course of 6ā7 weeks, with the patient receiving treatment 5 days per week. Radiation therapy schedules can sometimes be modified with hypofractionation and/or an abbreviated treatment course for elderly patients or those with a low performance status, while maintaining a similar level of toxicity and overall survival [15,16]. More aggressive approaches to irradiation using hyperfractionation schemes have not been shown to improve tumor control and, in some reports, have been associated with worse outcomes [14]. Other techniques to increase localized radiation doses to the tumor resection cavity, such as brachytherapy with permanent or temporary radioactive seeds, have also had disappointing results in controlled trials [17].
Stereotactic radiosurgery (SRS), using a linear accelerator-based system or Gamma Knife to deliver a single high-dose radiation fraction to a defined volume using stereotactic localization, is another method to boost radiation doses in the tumor bed of a newly diagnosed or recurrent high-grade astrocytoma [18,19]. Retrospective and single-armed uncontrolled trials suggest an improvement in local tumor control rates and survival when using either radiosurgical system. However, these results have not been confirmed in randomized, controlled trials, which have been performed within the Radiation Therapy Oncology Group (RTOG).
Chemotherapy of AA and GBMāHistorical Overview
In the 1960s and 1970s, surgical resection and postoperative external beam radiation therapy were established as the standard treatment approach for patients with AA and GBM [5,14]. However, in the late 1970s investigators began to evaluate chemotherapy as a potential treatment modality when it became obvious that surgical resection and radiation therapy were not curative in the vast majority of patients, and that other forms of treatment were desperately needed to improve survival [20ā24]. Over the next 15 years, numerous chemotherapy agents were tested, alone and in combination, for efficacy against newly diagnosed and recurrent high-grade gliomas. Single-agent nitrosoureas (i.e., BCNU) and nitrosourea-based combination regimens (i.e., PCV; procarbazine, CCNU, vincristine) were most effective (outlined in more detail below), although only to a modest degree. However, most of the literatures regarding chemotherapy for patients with AA and GBM have methodological deficiencies that impair the ability to draw firm conclusions [25]. Overall, there are very few well-controlled, prospective, randomized clinical trials of chemotherapy in patients with AA and GBM. The reasons for this remain somewhat obscure, but include the relative rarity of these patients, the initial lack of clinical infrastructure to perform large multiinstitutional clinical trials, the frequent presence of therapeutic nihilism in physicians caring for AA and GBM patients, and the fact that many of these patients were treated by different groups of subspecialty physicians (e.g., neurosurgeons, neuro-oncologists, radiation oncologists, medical oncologists) that have varying levels of interest in referral to clinical trials for chemotherapy. Owing to poor accrual and the small numbers of patients in many clinical trials, statistical power has been limited, further hampering the efforts to draw meaningful conclusions from the available data. Another set of problems that have diluted the quality of published results are the inclusion of...