Principles and Tenets of Radiation Treatment in Glioblastoma

Historical Context of Radiation Therapy and Dose

Historically, standard treatment for glioblastoma (GBM) was surgical resection alone. The first randomized trial to show a survival benefit with adjuvant radiation therapy (RT) was the Brain Tumor Study Group trial published in 1978, which showed a median survival of 37.5 weeks for RT alone, 25 weeks for adjuvant carmustine [1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU)] chemotherapy alone, and 17 weeks for supportive care without adjuvant treatment; combination of RT plus (+) BCNU yielded a survival of 40.5 weeks. In this study, whole-brain radiation therapy (WBRT) was delivered with parallel opposed fields to a dose of 50 to 60 Gy. The same group also conducted a dose-response analysis from less than 45 to 60 Gy and found improvement in median survival for doses of 50 to 60 Gy. A combined Eastern Cooperative Oncology Group (ECOG)/Radiation Therapy Oncology Group (RTOG) study in 1983 compared standard 60 Gy WBRT with 3 other arms: 60 Gy WBRT + 10 Gy partial-brain RT boost, 60 Gy WBRT + BCNU, and 60 Gy WBRT + lomustine and dacarbazine. This study showed no survival benefit with the 10 Gy boost and 60 Gy became the standard treatment dose with external beam RT (EBRT) for GBM.

Interstitial brachytherapy is a form of internal radiation that involves intraoperative placement of small radioactive sources into a tumor or resection cavity. Brachytherapy allows the delivery of high doses of radiation with significant dose fall-off at a short (brachy) distance to minimize damage to surrounding tissues. Early attempts at dose escalation were investigated using interstitial brachytherapy with iodine 125 (I-125) in 2 prospective randomized trials. The first randomized trial from Princess Margaret Hospital compared EBRT to 50 Gy in 2 Gy fractions versus (vs) EBRT 50 Gy + I-125 brachytherapy implant delivering an additional 60 Gy to the tumor or resection cavity. This study showed no survival benefit with the brachytherapy implant, yielding a median survival of 13.2 months in the standard arm vs 13.8 months in the brachytherapy arm ( P = .49). The Brain Tumor Cooperative Group went on to conduct the largest prospective randomized study with brachytherapy for malignant gliomas with 270 patients enrolled. Patients were assigned to either EBRT (60.2 Gy in 35 fractions) + BCNU or EBRT + BCNU + I-125 implant (60 Gy). Median survival was 68.1 weeks with I-125 compared with 58.8 weeks without I-125 ( P = .101). The lack of a statistically significant survival benefit despite large boost doses of brachytherapy, along with the logistical complexity, time factor, and operator dependence of the procedure, tempered the impetus for further investigation with brachytherapy. Enthusiasm for brachytherapy has further waned with advances in EBRT for dose escalation, including stereotactic radiosurgery as described later in this chapter as well as heavy particle RT such as proton therapy and carbon ion therapy.

With the advent of three-dimensional (3D) conformal RT (3DCRT), the dose could be further escalated to the tumor volume with EBRT. A study from the University of Michigan examined dose escalation up to 80 Gy without dose-limiting toxicity. Despite higher doses, 89% of patients developed an in-field recurrence. RTOG 8302 was a prospective phase I/II trial comparing dose escalation with hyperfractionation or accelerated hyperfractionation with BCNU. Hyperfractionation dosing included 64.8 Gy, 72 Gy, 76.8 Gy, and 81.6 Gy delivered twice daily in 1.2 Gy fractions, whereas accelerated hyperfractionation included 48 Gy and 54.4 Gy in 1.6 Gy twice-daily treatments. A preliminary report from this trial showed best survival in the 72 Gy hyperfractionation arm. Subsequently, a phase III study (RTOG 9006) examined 72 Gy hyperfractionation vs 60 Gy conventional fractionation; however, there was no survival benefit to hyperfractionation. Therefore, 60 Gy in 2 Gy daily fractions remains the standard of care with EBRT.

Current Standard of Care

The current standard of care for management of GBM is maximal surgical resection followed by concurrent chemoradiation therapy (chemoRT) with daily temozolomide (TMZ) to a radiation dose of 60 Gy, followed by further adjuvant TMZ. This regimen is based on level I evidence from the landmark European Organisation for Research and Treatment of Cancer (EORTC)–National Cancer Institute of Canada (NCIC) study published in 2005 by Stupp and colleagues, which showed a survival benefit to chemoRT with TMZ vs adjuvant RT alone. Before this study, adjuvant RT + BCNU was considered standard of care, but no randomized phase III trial had shown a statistically significant survival benefit with RT + BCNU compared with adjuvant RT alone for GBM. RT in the EORTC-NCIC study was delivered with 3DCRT to a dose of 60 Gy in 2 Gy fractions 5 days per week with a 2-cm to 3-cm margin around the gross tumor volume. For the chemoRT arm, TMZ was administered 7 days per week at a dose of 75 mg/m ² from the first day until the last day of RT. After a 4-week break, adjuvant TMZ dosed at 150 to 200 mg/m ² for 5 days was administered every 28 days for 6 cycles. Because TMZ can lead to lymphocytopenia, patients were administered prophylaxis against Pneumocystis carinii pneumonia with either pentamidine or trimethoprim-sulfamethoxazole. Median survival was 14.6 months with RT + TMZ and 12.1 months with RT alone, which translates to a 37% relative reduction in risk of death ( P <.001). Survival at 2 years was 26.5% for RT + TMZ and 10.4% with RT alone. Tumor progression was defined as an increase in tumor size by 25%, the appearance of a new lesion, or an increased need for corticosteroids. Median progression-free survival (PFS) was 6.9 months for RT + TMZ vs 5 months for RT alone ( P <.001). Only 8% of patients discontinued adjuvant TMZ because of toxic effects, and grade 3 or 4 toxicity was seen in 7% of patients in the RT + TMZ arm.

An update of the EORTC-NCIC trial reported that the long-term survival advantage with TMZ persists at 5 years follow-up. Overall survival (OS) comparing RT + TMZ vs RT alone was 16% vs 4.4% at 3 years, 12.1% vs 3% at 4 years, and 9.8% vs 1.9% at 5 years, respectively (hazard ratio, 0.56; 95% confidence interval [CI], 0.47–0.66; P <.0001). Patients who had gross total resection survived longer than those with subtotal resection. The worst outcome was in patients with unresectable tumors who had undergone biopsy only. Promotor methylation of O6-methylguanine-DNA methyltransferase (MGMT) was the strongest prognostic factor and predictor for survival with TMZ, as discussed in further detail later in this chapter.

Immobilization

To minimize daily setup errors and intrafractional patient movement, creating a reproducible immobilization device is paramount to delivering an accurate RT plan. Patients should be simulated supine on a head cup or pad, with a thermoplastic mask that conforms to the patient’s face for immobilization ( Fig. 8.1 ). This position confers a daily setup error of 3 mm, which is the current standard for clinical target volume (CTV) to planning target volume (PTV) expansion when creating RT volumes. Newer technologies include an open-faced mask that uses an optical surface tracking system via camera pods to capture facial features for tracking to ensure correct patient setup. The benefit of this newer system is that it circumvents daily x-ray radiation exposure for patient alignment and potentially mitigates patient discomfort and claustrophobia compared with a standard thermoplastic mask. This technology also allows radiation oncologists to track intrafractional movement with the ability to halt RT via real-time feedback if the patient moves past a threshold (eg, 3 mm) during treatment. Once the patient is immobilized, a computed tomography (CT) simulation scan is performed using 1-mm to 3-mm slice thickness from vertex to below the skull base for treatment planning.

Target Volumes

A postoperative MRI scan including T1-weighted postcontrast (T1C+) MRI and T2-weighted fluid-attenuated inversion recovery (FLAIR) MRI should be fused with the CT simulation scan to generate target volumes for radiation treatment planning. A postoperative MRI within 72 hours after surgery allows for the best assessment of the extent of residual tumor before blood and postoperative edema cloud the clinical picture. MRI performed <2 weeks before RT is ideal because further edema may cause midline shift. There is also the potential for tumor regrowth from the time of surgery to RT. Although functional imaging modalities such as magnetic resonance spectroscopy (MRS) and dynamic contrast-enhanced MRI may potentially add information regarding target volume contouring and treatment planning, these modalities have not been validated and are still considered investigational. There are some differences in target volume delineation between the EORTC and the RTOG, as detailed in ( Table 8.1 ). The main difference is that the RTOG treats the FLAIR signal hyperintensity with a margin to account for peritumoral spread from edema followed by a cone-down boost to the T1C+ MRI enhancement, whereas the EORTC treats only the T1C+ MRI volume without a boost. Regardless, both the RTOG and EORTC advocate a 2-cm volumetric 3D expansion around the gross tumor volume (GTV) visualized on T1C+ MRI to create the CTV, with a reduction in margins to respect anatomic barriers, including the skull, ventricles, falx cerebri, tentorium cerebelli, optic chiasm/nerve, and brainstem. The CTV margins are based on historical studies showing that approximately 80% of recurrences are within a 2-cm margin of the enhancement seen on T1C+ MRI scans. An additional 3-mm to 5-mm margin is added to the CTV to account for daily setup error to create the PTV. An example of target volume contouring based on RTOG guidelines is shown in Fig. 8.2 . RTOG 0525 and CENTRIC clinical trials allowed for contouring based on either guideline and showed no difference in PFS or OS when comparing the 2 guidelines for target volumes. Retrospective studies comparing the EORTC and RTOG contouring guidelines also showed no difference in tumor recurrence. In the United States, most radiation oncologists follow the RTOG guidelines, although the EORTC margins may lead to reduced toxicity with no proven disadvantage in local control or survival.

Organs at Risk

Organs at risk (OARs) to be contoured along with tolerance doses and toxicity are summarized in Table 8.2 . Doses to OARs should be evaluated by a dose-volume histogram (DVH). Sometimes the dosimetric goals of the OARs may not be achievable because of the location and size of the PTV. In these scenarios, the radiation oncologist must make a clinical decision regarding the risks of reducing PTV coverage against the benefits of avoiding potential radiation damage to OARs. The most critical organs to avoid exceeding the tolerance dose are the brainstem, optic chiasm, and optic nerves, because of the potentially severe consequences of radiation-induced injury to these structures. Some radiation oncologists incorporate a planning risk volume (PRV) around the OARs, which is congruous to the CTV → PTV expansion, to also account for variability in daily setup for the OARs during fractionated RT.

Treatment Planning and Delivery

In the past, 3DCRT was used for partial-brain treatment planning for GBM. However, with modern treatment techniques such as intensity-modulated RT (IMRT) and volumetric intensity-modulated arc therapy (VMAT), which are now available at most radiation centers, a more conformal treatment plan can be created to deliver the prescribed treatment dose while minimizing dose to OARs. Ideally, the treatment planning goals should be the following: 95% of the PTV should receive 100% of the prescription dose (D95 = 100%) and 100% of the CTV and GTV should receive 100% of the prescription dose. Maximum plan dose (hot spots) should be ≤115% of the prescription dose. Daily image-guided RT with cone-beam CT should be aligned to the skull to maintain treatment accuracy. A treatment plan with isodose lines and DVH for a patient with GBM is shown in Fig. 8.3 .