Radiation Therapy in High-Grade Gliomas
- 315 Downloads
High grade gliomas (HGG) account for the majority of primary malignant brain tumors in adults with increasing incidence in age, peaking in the 75–84 age group. Glioblastoma (GB) is the most frequently occurring subtype, followed by anaplastic astrocytoma (AA), anaplastic oligodendroglioma (AO), and mixed anaplastic oligoastrocytoma (AOA). Since 2016 a new WHO classification has radically changed our understanding of gliomas and introduced a molecular classification system allowing for future, more personalized, treatment approaches.
As the incidence of HGG increases with advancing age, selection of optimal treatment is complicated by preexisting comorbidities and deteriorating performance status. HGG are aggressive tumors which are best managed with a combined modality approach. Maximal surgical resection is associated with improved survival and favorable neurological improved outcome. However, due to the infiltrative nature of HGG, a complete microscopic clearance is rarely possible.
Radiotherapy post surgery has been the mainstay of treatment for HGG for several decades. Involved field radiotherapy is now the standard approach used to maximize coverage of infiltrative tumor cells, while limiting the volume of normal brain tissue irradiated. Dose escalation above 60Gy, hypo- or hyper-fractionation has not shown any additional benefit. The introduction of concomitant and adjuvant temozolomide to radical radiotherapy has provided a modest but significant improvement in overall survival for patients with GB.
Despite optimal treatment, the prognosis for HGG remains poor with limited treatment options at recurrence. Additional molecular pathways identified in the development of HGG are now the target for novel therapies, which may potentially improve the prognosis of what remains to be a challenging and highly fatal disease.
KeywordsRadiotherapy High grade glioma Malignant glioma Personalised medicine Target volume definition
Gliomas account for the majority of primary malignant brain tumors in adults, with high grade gliomas (HGG) (or equally referred to as malignant gliomas) representing up to 75% of all diagnosed gliomas of the central nervous system. The annual incidence ranges from 3 to 5 per 100,000 with a slight predominance in males. The peak incidence of HGG is in the fifth and sixth decades of life, although HGG can develop in all age groups. (Dolecek et al. 2012) HGG comprise of World Health Organisation (WHO) Grade III tumors (anaplastic astrocytoma (AA), anaplastic oligodendroglioma (AO), mixed anaplastic oligoastrocytoma (AOA), and WHO Grade IV tumors (glioblastoma (GB)) (Louis et al. 2016). GB has traditionally been referred to as glioblastoma multiforme (GBM) due to its multicolored appearance on H&E staining, but this nomenclature has been modified in the most recent WHO classification system. GB is the commonest form of high grade glioma, accounting for 60–70% of all diagnosed cases. Anaplastic astrocytoma accounts for approximately 10–15% of cases, with anaplastic oligodendrogliomas and anaplastic oligoastrocytomas occurring less frequently, in up to 10% of cases (Sathornsumetee et al. 2007). Other, rare malignant glioma subtypes include anaplastic ganglioglioma, anaplastic pilocytic astrocytoma, anaplastic pleomorphic xanthoastrocytoma (all classified as WHO III°), giant cell and small cell glioblastoma, epithelioid glioblastoma, and gliosarcoma (WHO IV°). Anaplastic astrocytoma and glioblastoma increase in incidence with age, peaking in the 75–84 age group, while oligodendrogliomas and oligoastrocytomas are most commonly found in the 35–44 age group.
The incidence of HGG has been rising in recent decades, especially among the elderly population. This is most likely a result of improved and more widely available diagnostic imaging, along with safer and more frequent surgical verification of the diagnosis even in the elderly population, rather than a genuine increase in incidence.
The underlying cause for the development of HGGs in the majority of cases has not been identified. Exposure to ionizing radiation remains the only established risk factor for the development of malignant glioma (Fisher et al. 2007). Evidence for a potential link with prior head injury, dietary nitrites, and occupational exposure remains inconclusive. There has been recent concern regarding the potential increased risk of developing gliomas with the use of mobile (cellular) phones; however, a number of large scale population-based studies have not been able to demonstrate a link (Lahkola et al. 2007).
Despite optimal treatment, the median survival for GB as a whole remains poor with an expected median survival in the range of 12 to 18 months, a 2-year survival of 12 to 26% and an approximately 10% 5-year survival rate when treated with radical intend (Stupp et al. 2009). Patients with anaplastic astrocytoma have a better median survival ranging from around between 2 and 3 years (Wick et al. 2009). Prognosis is best for patients with anaplastic oligodendroglioma, with an expected median survival of approximately 12–15 years (Cairncross et al. 2013).
A number of rare genetic disorders predispose to the development of HGG. The majority of these are inherited defects in the regulation of cell proliferation and apoptosis, caused by germline mutations, such as Neurofibromatosis type 1, Li-Fraumeni syndrome, and Cowden’s syndrome. Lynch syndrome, Turcot’s syndrome, Tuberous Sclerosis, and von Hippel-Lindau disease are also, but less frequently, associated with primary gliomas (Farrell and Plotkin 2007). Approximately 5% of patients with malignant gliomas have a family history of the disease. However, most familial cases have no identifiable underlying genetic cause. The recently established international consortium, GLIOGENE, has been tasked with studying the genetic bases of familial gliomas by screening 15,000 individuals worldwide and may provide some insight into the genetic links (Malmer et al. 2007).
On the basis of genetic differences, glioblastomas are separated into two molecularly distinct subtypes: primary (de novo) and secondary GB (Furnari et al. 2007). Primary GB are characterized by epidermal growth factor receptor (EGFR) amplification and mutations, loss of heterozygosity of chromosome 10q, p16 deletion, and deletion of phosphatase and tensin homologue (PTEN) on chromosome 10. Secondary GBs arise from preexisting low grade astrocytomas over a period of several years. They are characterized by the presence of an isocitrate dehydrogenase gene (IDH 1/2) mutation, loss of p53 tumor suppressor gene, overexpression of platelet derived growth factor receptor (PDGFR), abnormalities in p16 and Rb pathways, and loss of heterozygosity of chromosome 10q. Although there are clear genetic differences in primary and secondary GBM, morphologically they remain indistinguishable.
Concomitant deletion of chromosomes 1p and 19q (1p/19q co-deletion) is characteristic of oligodendroglioma and absent in astrocytoma. In the early phase of oligodendroglioma pathogenesis, an unbalanced translocation between chromosome 1 and 19 with loss of the derivative t(1p;19q) results in loss of heterozygosity (LOH). Chromosome 1p/19q co-deletion is now pathognomonic for a diagnosis of oligodendroglioma and seen in up to 90% of patients with high grade oligodendroglioma and is associated with improved survival in patients treated with chemotherapy and/or radiotherapy (Jenkins et al. 2006).
The tumors with a 1p/19q co-deletion also have mutations in the isocitrate dehydrogenase gene (IDH). The gene encoding for isocitrate dehydrogenase 1 (IDH1) is mutated in astrocytomas, oligodendrogliomas, and mixed oligoastrocytomas of WHO grades II and III, as well as secondary glioblastomas (Louis et al. 2016). Loss of function of IDH1 may have an indirect oncogenic effect through activation of the hypoxia-inducible factor pathway. This is associated with metabolic adaptation of tumors to anaerobic growth and angiogenesis (Belozerov and Van Meir 2006). Recurrent mutations in the active site of IDH1 and IDH2 have been found in up to 70% of low grade gliomas and 10% of GB patients (Louis et al. 2016).
Additional molecular pathways involved in the process of developing gliomas include the epidermal growth factor receptor (EGFR) pathway; the retinoblastoma (Rb) tumor-suppressor pathway; the phosphatidyl-inositol 3-kinase (PI3K) pathway; and the DNA repair pathways involving the methylguanine methyltransferase (MGMT) protein and mismatch repair (MMR) enzymes.
DNA methylation of the O6-methylguanine-DNA methyl transferase (MGMT) gene promotor region results in epigenetic modification. Methylation of the gene promotor can reduce expression of the DNA repair gene. MGMT is an important repair enzyme that has been associated with increased resistance to alkylating chemotherapy agents. Therefore, methylation of the gene is predictive for benefit from alkylating chemotherapy in GB.
Methylation of MGMT promoter is highly associated with LOH chromosome 1p/19q and IDH mutations (Sanson et al. 2009). A recent study published by the German Neuro-Oncology Working Group (NOA) reported that MGMT promoter methylation is a predictive biomarker for benefit from alkylating agents in patients with IDH1-wild-type glioma, while it is of no significant benefit for IDH1-mutant disease (Wick et al. 2013). Studies have also found a strong link between IDH mutations and a genome-wide glioma cytosine–phosphate–guanine island methylator phenotype (G-CIMP) across all glioma subtypes (Noushmehr et al. 2010)
The most common defects in growth factor signaling involve EGFR and PDGFR. PDGF signaling is a key regulator of glial development and may be exploited as a therapeutic target in the future. Amplification of EGFR is mainly seen in primary glioblastoma and is present in approximately 40–50% of tumors. Almost half of tumors with EGFR amplification express an autophosphorylated variant of EGFR known as EGFR VIII.
Mutations in the BRAF V600E oncogene have been identified in a small group of pediatric and adult glial-based tumors and offer a targeted therapeutic option with currently available drug agents. This mutation commonly occurs in WHO Grade II and III pleomorphic xanthoastrocytomas (PXA) and is also a frequently identified mutation in epithelioid glioblastoma (Kleinschmidt-DeMasters et al. 2013).
Programmed cell death ligand 1 expression (PD-L1) has been found in up to 88% of newly diagnosed and 72% of recurrent glioblastoma histological specimens (Avril et al. 2010). The highest rates of PD-L1 expression have been seen in mesenchymal GB subtypes and are associated with poorer survival outcomes. PD-L1 expression has not yet been found to have any prognostic value in glioma; however, it offers a potential therapeutic target with immune checkpoint inhibitors.
Presenting symptoms of HGG are usually caused by the mass effect exhibited by the tumor, peritumoral edema, secondary raised intracranial pressure, and direct infiltration/compression of adjacent functional critical brain structures. The individual symptoms or clinical signs are more accurately appreciated by the location and rate of growth of the tumor. Rapidly progressing tumors located along the ventricular system or within eloquent areas of the cerebral cortex may become clinically apparent after only a small amount of growth. Tumors arising in less eloquent parts of the brain frequently only present after a substantial growth period.
Patients will present with a variety of symptoms many of which initially are frequently nonspecific, including cognitive impairment, behavioral change, early morning nausea/vomiting, diplopia, papilloedema, persistent headaches, and seizures. The new onset of seizures in patients over the age of 40 should be considered as highly suspicious of a CNS malignancy unless proven otherwise. Symptoms related to direct infiltration or compression of adjacent structures may include focal motor deficit, dysphasia, pyramidal tract findings, or cerebellar signs (Wen et al. 2006).
Although HGG are invasive tumors, dissemination of disease is overwhelmingly limited to the central nervous system; therefore, regular extra cranial systemic staging is not required. Staging of HGG is based on magnetic resonance imaging (MRI) and computerized tomography (CT). In the absence of suspicious clinical symptoms, suggestive of metastatic spread cerebrospinal fluid (CSF) analysis and staging of the spine is generally not indicated.
Contrast enhanced MRI has a higher resolution and sensitivity compared to CT and is the preferred imaging modality from a diagnostic perspective. In addition, it can provide accurate three-dimensional reconstructed images valuable in the preoperative setting.
Advanced MRI techniques, such as diffusion-weighted, perfusion-weighted and MR spectroscopy, as well as different PET modalities, can provide valuable additional information on cellularity, tissue structure, metabolism, and vascularity. These techniques can furthermore aid diagnosis, provide noninvasive prognostic biomarkers, help guide an appropriate and safe biopsy trajectory, assist treatment planning, and monitor clinical outcome. However, these techniques are currently not incorporated into routine clinical practice, and robust data obtained from multicenter clinical trials are yet required to further define their role.
The acute management of high grade glioma is focused on the management of direct tumor-related symptoms, especially raised intracranial pressure (ICP) and mass effect. Raised ICP is a medical emergency, requiring immediate intervention. Standard initial treatment includes high dose corticosteroids and/or the use of osmotic agents. Surgical decompression in the acute setting is rarely required and is reserved for an agreed and planned primary management strategy.
Corticosteroids, such as Dexamethasone, are frequently used to treat tumor-associated edema. Doses of dexamethasone 8–16 mg/day are effective for rapid reduction of edema and improvement of clinical symptoms. Symptom-driven regular tapering and early discontinuation of steroids is recommended to avoid severe toxicity associated with prolonged use (e.g., cushingoid appearances, proximal myopathy, steroid-induced diabetes mellitus). Continuation of steroid therapy after tumor resection or for prophylaxis during radiotherapy in asymptomatic patients is not recommended.
Approximately 25–40% of patients with HGG will present with or develop tumor-related seizures. This is thought to occur more frequently in patients whose tumors have an oligodendroglial origin (You et al. 2012). Antiepileptic therapy is indicated for patients presenting with seizures. First generation antiepileptic drugs that are strong inducers of hepatic cytochrome P450 enzymes, such as phenytoin and carbamazepine, can interact with commonly used chemotherapy agents. For this reason, anticonvulsants such as Levetiracetam or Lamotrigine are currently preferred. Prophylactic use of anticonvulsants outside the peri-operative phase in patients who have never had a seizure is not indicated. (Weller et al. 2012)
Patients with malignant glioma are at an increased risk of venous thromboembolism (VTE). This is due to a tumor-induced hypercoagulable state, immobility as a result of neurological deficits, or prolonged steroid use (Perry 2012). Prophylactic anticoagulation is not recommended. There is no contraindication to the use of therapeutic anticoagulation in patients with confirmed thrombosis. The risk of intralesional hemorrhage associated with anticoagulation therapy in patients with glioma is low, unless a patient has an existing intracerebral hemorrhage or other preexisting contraindications.
The biological hallmark for malignant glioma is local invasion of surrounding tissue. Traditionally, gliomas have been stratified into grades based on their microscopic histological appearance.
Astrocytic tumors comprise of elongated or irregular cells, hyperchromatic nuclei, and glial fibrillary acidic protein (GFAP) positive cytoplasm. Oligodendrogliomas have rounded nuclei, perinuclear halos, calcification, and branching blood vessels. Nuclear atypia and increased mitotic activity is characteristic of anaplastic, grade III tumors. Microvascular proliferation and necrosis are the hallmark of grade IV tumors.
The 2016 WHO classification system for CNS tumors upgraded the classification of glioma by incorporating specific genetic and molecular parameters to help define the separate tumor entities. This has specifically impacted the classification of grade III astrocytoma, oligoastrocytoma, and oligodendroglioma. They are now stratified on the basis of growth pattern, IDH mutation status, 1p/19q co-deletion status, and ATRX mutation status (Louis et al. 2016).
Mutations in IDH1/2 are a defining feature of grade II and III diffuse astrocytic and oligodendroglial tumors. The majority of glioblastomas are IDH wild type. Immunohistochemical staining for IDH mutations or DNA sequencing should be carried out on all diffuse glioma biopsy or surgical specimens, or in GB patients below the age of 55 years (Horbinski et al. 2009; Hartmann et al. 2009).
1p/19q co-deletion is a defining feature of oligodendroglioma and is co-expressed with IDH1/2 mutations. Testing for co-deletion status through fluorescence in situ hybridization (FISH) should be performed on all tumors displaying oligodendroglial differentiation. Oligodendroglial tumors have a substantially better prognosis in comparison to primary astrocytic tumors.
IDH mutant diffuse gliomas with loss of ATRX and TP53 mutation are classified as diffuse astrocytoma. IDH mutant with 1p/19q co-deletion is oligodendroglioma. IDH wild-type grade III gliomas are classified as diffuse astrocytoma, IDH wild-type glioma, or oligodendroglioma NOS (not-otherwise-specified).
Patient assessment and treatment plan should be managed by a specialist multidisciplinary team comprising of neurosurgeons, oncologists, neuropathologists, neuroradiologists, and neurologists. Initial management strategy should consider patient performance status, comorbidities, presenting symptoms, neurological function, radiological findings and, if appropriate, response to high dose steroids.
Histopathological diagnosis for patients with suspected high grade glioma is essential. A stereotactic biopsy can be offered to virtually every patient regardless of location of the primary tumor. CT or MRI imaging and frameless stereotactic devices allow neurosurgeons to perform an accurate guidance system to allow for a safe intervention. This procedure is associated with a high rate of obtaining sufficient tissue for a definitive pathological and molecular diagnosis (Paleologos et al. 2001).
The primary goal of surgery in patients with high-grade glioma is both to confirm the histopathological diagnosis and achieve a maximal safe resection in the form of a subtotal or, ideally, gross total resection as this is associated with improved outcomes (Noorbakhsh et al. 2014). This aim however must be balanced with the preservation of neurological function, particularly if the tumor is located in dominant or eloquent areas of the brain (Lacroix et al. 2001). A number of intraoperative techniques are used to improve the extent of surgical resection, while minimizing potential damage to normal brain. These techniques include, among others, pre- and intraoperative ultrasound or MRI, cortical mapping/stimulation, awake craniotomies, and the use of 5-aminolevulinic acid (5-ALA) for fluorescent marking of tumors (Senft et al. 2011; Stummer et al. 2006).
The extent of resection (EOR) has been demonstrated to have a prognostic impact and thus an attempted macroscopic complete resection without unacceptable damage to adjacent functional brain tissue is the primary goal at surgery (Grabowski et al. 2014). A recent Surveillance, Epidemiology and End Results (SEER) registry study noted that gross macroscopic resection was associated with an improved median overall survival of two to three months across all age groups in comparison to a biopsy (Noorbakhsh et al. 2014).
Seminal studies have noted a significant survival advantage in association with surgical resections of 98% or greater of tumor volume (median survival 13 months, 95% confidence interval [CI] 11.4–14.6 months), compared with 8.8 months (95% CI 7.4–10.2 months; p < 0.0001) for resections of less than 98% (Grabowski et al. 2014). However, the threshold value for the extent of resection correlating with improved survival has long remained a source of debate. There is now mounting evidence that total or near total resections of greater than 90 to 95% of tumor volume is the likely threshold to confer to a survival benefit (Hardesty and Sanai 2012). There is also evidence suggesting a progression-free survival benefit with subtotal resections of as low as 78% (Sanai et al. 2011).
In addition to diagnostic and prognostic benefits, maximal resection allows for symptomatic relief and neurological improvement. This may also facilitate speedy tapering and discontinuation of steroids, therefore reducing the risk of steroid-induced complications.
Role of Chemotherapy
The choice of first line chemotherapy differs depending on histological diagnosis.
The use of systemic chemotherapy for GBM has been shown to add a small, but statistically significant improvement to overall survival (Stewart and Group GM 2002). The use of single or multidrug regimens containing nitrosureas, given in combination with radiotherapy has traditionally resulted in conflicting results. A meta-analysis found the addition of nitrosurea-based chemotherapy to radiotherapy to be beneficial, with an improvement in 1-year survival rates of 6%. A separate MRC study did not find any survival benefit with the addition of multiagent PCV chemotherapy (procarbazine, lomustine [CCNU], vincristine) (Medical Research Council Brain Tumor Working Party 2001).
In recent years, the addition of concomitant and adjuvant temozolomide (TMZ) to radiotherapy for newly diagnosed GB has been shown to provide a clinically meaningful and statistically significant survival benefit for patients with WHO performance status of 0 and 1.
The pivotal study investigating the use of TMZ with radiotherapy for GB was reported by Stupp et al. in 2005, and the improved outcome with concomitant chemoradiation was confirmed with longer follow-up (Stupp et al. 2009). The two-year survival rate was 26.5% with radiotherapy plus temozolomide versus 10.4% with radiotherapy alone. This approach has now been accepted as the international standard of care for patients with a good performance status.
A subset analysis of the Stupp data revealed that benefit from TMZ was most pronounced in GB patients with MGMT promoter methylation. Two-year survival was 49% for combination therapy and 24% for radiotherapy alone in patients with MGMT methylation.
There is currently no data to support the use of temozolomide beyond 6 cycles in the adjuvant phase for GB (Gilbert et al. 2013).
The addition of Bevacizumab (a monoclonal anti-VEGF antibody) to radiotherapy and temozolomide has been shown to have a modest effect on progression free survival, delaying progression by three to four months. There has been no proven improvement in overall survival; however. the drug has a role to play in the management of recurrent disease (Chamberlain 2011).
Local chemotherapy in the form of carmustine (BCNU) wafers placed in the surgical cavity intra-operatively, followed by radiotherapy, has been shown to marginally improve median survival compared to radiotherapy alone in patients with tumor resection of greater than 90% and no surgical opening of the ventricular spaces (Westphal et al. 2003). There is no consensus on the use of this treatment approach in the era of radiotherapy and concomitant/adjuvant temozolomide as the survival benefit was not seen when the data were stratified by GB diagnosis only (Dixit et al. 2011).
A number of other novel agents such as Cilengitide, a selective integrin inhibitor, have been studied in combination with TMZ and radiotherapy, with no demonstrable improvement in overall survival (Stupp et al. 2014).
Anaplastic Astrocytoma (WHO Grade III)
Findings from the German neuro-oncology group (NOA)-04 trial have shown that chemotherapy alone (either PCV or temozolomide) is as effective as radiotherapy alone in terms of progression-free survival, with similar outcomes in overall survival (Wick et al. 2009). Temozolomide is frequently used as the first line chemotherapy of choice due to its favorable toxicity profile in comparison to PCV.
Preliminary results from the CATNON trial, a randomized controlled phase III trial of 748 patients with anaplastic gliomas without 1p19q co-deletion, have suggested a significant improved overall survival with the use of concomitant and adjuvant temozolomide in addition to-radiotherapy, when compared to primary radiotherapy alone (HR 0.65, 95% CI 0.45–0.93) (Van Den Bent et al. 2017).
Anaplastic Oligodendroglioma and Oligoastrocytoma (WHO Grade III)
Oligodendroglial tumors are known for their enhanced sensitivity to alkylating agents and temozolomide. For patients with 1p/19q co-deleted tumors, PCV given before or immediately after radiotherapy does carry a statistically significant survival benefit, although this benefit only manifests after 6 years of follow-up (van den Bent et al. 2013).
Role of Radiotherapy
Despite advances in surgical resection techniques and a higher rate of macroscopically complete resections, HGG cases are still virtually always associated with the presence of macroscopic or microscopic residual disease, due to their diffuse infiltrative nature.
The survival advantage of adjuvant radiotherapy for HGG patients was first demonstrated by the NCI Brain Tumor Study Group studies published in the 1970s (Walker et al. 1978). Patients were initially treated with whole brain radiotherapy (WBRT) using parallel opposed beams. Over the decades, observational studies noted that recurrences following WBRT overwhelmingly occurred within 1-2 cm of the original contrast enhancing tumor site (up to 90% of cases) (Wallner et al. 1989; Hochberg and Pruitt 1980). Additionally, fewer than 10% of HGG recurrences are multifocal. Subsequent advances in imaging and radiotherapy techniques have allowed for dose escalation within tumor sites, while minimizing the irradiated volume of normal brain tissue. As a consequence, involved field radiotherapy has become the accepted standard of care worldwide. There is remaining controversy regarding how partial volume radiotherapy is best delivered in order to ensure that the infiltrating tumor cells lying beyond the radiologically visible tumor edge are adequately covered within the radiotherapy field, thus optimizing local control.
In North America, a two-phase technique is commonly used aimed at treating the radiographically apparent gross tumor and peritumoral edema in the first phase, followed by targeted boost to the area of residual contrast enhancement/tumor bed. Most European centers use a single phase primary target volume definition which correlates in general to the boost phase of the American radiotherapists, as recommended by a European multinational working group (Laperriere et al. 2002; Niyazi et al. 2016).
Radiotherapy Dose and Fractionation
The radiation tolerance of normal brain parenchyma and vasculature are dose limiting factors for external beam radiotherapy. The risks of acute and long-term toxicity increase with escalating radiation dose. Risk of severe toxicity increases with greater size of the planning target volume (PTV) and proximity of the PTV to critical organs at risk. The standard dose for radiotherapy is 60 Gy in 1.8–2 Gy per fraction for GB and 54–60 Gy in 1.8–2 Gy fractions for grade III AOs and AAs. Dose escalation beyond 60 Gy, as well as alternative radiotherapy approaches such as accelerated hyperfractionation, hypofractionation, brachytherapy, or radiosurgery, has not shown a survival benefit but can add to radiation associated toxicity.
The exception from this rule is the radiotherapy fractionation for HGG in the elderly population who are not deemed fit for radical chemoradiation or patients with poor performance status where short-term alleviation of symptoms is the primary aim. Hypofractioned radiotherapy is considered standard of care in this situation.
Target Volume Delineation
Radiotherapy Tumor Target Volumes
Specific normal tissue structures must be defined and identified as organs at risk (OARs). Small OAR should be given a set-up error margin to create a planning at risk volume (PRV). A radiotherapy technique that minimizes the amount of normal brain tissue and OARs exposed to high dose irradiation should be used. Suitable techniques include 3D conformal RT, fixed field intensity modulated radiotherapy (IMRT), and dynamic arc IMRT. 6MV photons are typically used to deliver RT. Dose variations across the PTV should be kept within +7% and – 5% of the prescription isodose according to ICRU recommendations (ICRU report 50, 62, and 83).
The radiotherapy planning CT scan is co-registered with a volumetric (3D) MRI scan within the RT treatment planning system (TPS). Three-dimensional scans (CT and/− MRI) should be acquired with the patient in the radiotherapy treatment position with a slice thickness of 2.5 mm or less. Ideally, the MRI should be obtained as close as possible to the time of radiotherapy, ideally within 2 weeks. The pre- and post-op MRI images should also be co-registered with the planning CT within the TPS at the time of contouring. If MRI is contraindicated, a contrast-enhanced CT scan should be performed.
Beam arrangements and dosimetry are determined by 3D treatment planning software. All target volumes (GTV, CTV, PTV, and OARs) should be delineated and reconstructed in 3D.
There are two main approaches toward tumor target volume definition for GBM between North America and Europe as mentioned above, although there is no demonstrated survival benefit to either approach (Zhao et al. 2016).
American clinicians prefer a two-phased approach (shrinking field technique). GTV1 is defined as the volume encompassing the contrast enhancing tumor tissue and all high signal changes as seen on T2 weighted imaging. GTV2 is defined as the volume encompassing only residual contrast enhancing tumor elements/tumor bed, as seen on postoperative MRI. GTV1 is expanded by 1.5-2 cm to form CTV1. GTV2 is expanded by 1 cm to form CTV2. Bone and dura act as natural barriers to disease infiltration; therefore, following manual expansion, CTV1 and CTV2 are manually edited to reside within the skull vault and to avoid crossing dural planes, unless there is definite radiological evidence of disease. The CTVs may also be manually modified to avoid the CTV to unnecessary cross the midline into areas of brain not at risk for recurrence and to avoid adjacent critical structures, e.g., the optic apparatus, pituitary, and brainstem. An isotropic margin is applied to the CTV to form the PTV. This volume must not be manually adapted. The margin applied is dependent on the patient positioning and immobilization systems used and is therefore institution dependent. In practice, the CTV-to-PTV margin added will typically be between 0.3 and 0.5 cm. In this two-phase approach, the typical PTV1 prescription dose is 46 Gy, and an additional dose of 14 Gy is prescribed to ensure that the final PTV2 does reaches 60 Gy.
European centers generally use a single-phase approach, as detailed in the ESTRO-ARCOP target volume delineation guideline (Niyazi et al. 2016). GTV is defined as the enhancing tumor, as visualized on gadolinium contrast enhanced T1-weighted MRI. The primary tumor volume (if no surgery conducted) or the entire surgical cavity is delineated. The GTV is volumetrically expanded by 1.5-2 cm and manually edited off anatomical boundaries to form the CTV identical to the American approach. The PTV is generated from the CTV by the addition of an isotropic margin according to departmental policy (again, usually 0.3–0.5 cm). The prescribed PTV dose is 60 Gy in conventional fractionation.
For macroscopically completely or nearly resected tumors, the GTV is delineated on the postoperative MRI T1 weighted images. The GTV includes the entire surgical tumor bed and all contrast-enhancing areas consistent with residual disease. The use of post-op diffusion weighted imaging (DWI) may help distinguish postoperative changes from residual disease.
Limiting margins: Recent advances in radiotherapy technique have prompted research into the safety of reducing the GTV-to-CTV margin for HGG. Margins as small as 0.5 to 1.0 cm have been studied, with no apparent increase in the local recurrence rate (Paulsson et al. 2014).
Organs at Risk (OARs)
The following organs at risk (OARs) are frequently defined on treatment planning systems: optic nerves (left and right), optic chiasm, eyes (left and right), lenses (left and right), cochlea (left and right), and whole brain and brainstem. Additional OARS for consideration include lacrimal glands, pituitary gland, and hypothalamus. The hippocampus can also be contoured if the tumor is in a location that will allow for hippocampal sparing without compromising the PTV dose. Organ tolerance doses should be defined as per QUANTEC data.
Optic Nerves and Chiasm
The optic nerves and optic chiasm may be contoured on the RT planning CT or on a co-registered T1-weighted post contrast MRI. The optic chiasm is most readily identified on coronal images prior to outlining on axial images on the TPS. The right and left optic nerves are contoured separately, starting from the posterior globes to their exit from the optic canals. The optic chiasm is contoured as a single structure consisting of pre- and postchiasmal component and a central body. The prechiasmal component starts at the optic canals and terminate at the anterior body of the chiasm; the postchiasmal component starts from the posterior body of the chiasm and run posteriorly prior to entering the thalami. The incidence of radiation-induced optic neuropathy (RION) is uncommon with a Dmax <55 Gy at fraction sizes of <2 Gy. The RION risk increases to 3%–7% for doses in the region of 55–60 Gy with fraction sizes of 1.8–2.0 Gy (Mayo et al. 2010a).
Each eye is separately contoured on the RT planning CT from the most superior to inferior aspect. The whole outside of the globe should be contoured to include sclera and cornea. Effort should be made to avoid direct treatment of the anterior chamber of the eye to protect the lenses and cornea. Dose to the entire eye should be kept to the minimum possible, without compromising PTV coverage. DMax for macula should be 45 Gy or less.
The brainstem is contoured directly on the RT planning CT. The volume includes the midbrain, pons, and medulla. The cranial boundary is inferior to the third ventricle and optic tracts. The caudal boundary of the brainstem is taken as the foramen magnum. The entire brainstem can be treated to 54 Gy using fraction sizes of <= 2 Gy with a low risk of irreversible neurological toxicity. Smaller volumes of the brainstem (1–10 cc) may be irradiated to a maximum dose of 59–60 Gy at dose fractions of <= 2 Gy, and the risk of neurological toxicity increases markedly at doses beyond 64 Gy (Mayo et al. 2010b).
Each cochlea is contoured on the RT planning CT, viewed using bone-windows, as a circular structure within the petrous portion of the temporal bone anterior to internal acoustic meatus (IAM). The contour should appear on at least two successive axial CT slices. The aim is for the cochlea dose to be no more than D50% < 35 Gy for hearing preservation (Bhandare et al. 2010).
The hippocampal structures are best contoured on a T1-weighted MRI co-registered with the RT planning CT, according to the Radiation Therapy Oncology Group (RTOG) hippocampal outlining atlas (Gondi et al. 2010). Retrospective data from the Childhood Cancer Survivor Study (CCSS) demonstrate an increased risk for memory difficulties with RT doses above ≥30 Gy to the temporal lobe region (Armstrong et al. 2010). Effort should be made to minimize dose to the hippocampi but without compromise of PTV coverage, which always takes priority.
The pituitary may be contoured on the RT planning CT. Retrospective data from a cohort of childhood cancer survivors treated with cranial radiotherapy for a variety of diagnoses found that, compared with total RT doses <22 Gy, doses of 22 Gy to 29.9 Gy were associated with Growth Hormone (GH) deficiency; doses ≥22 Gy were associated with Luteinizing hormone and follicle-stimulating hormone (LH and FSH) deficiencies; and doses ≥30 Gy were associated with thyroid-stimulating hormone (TSH) deficiency and adrenocorticotropic hormone (ACTH) deficiency (Chemaitilly et al. 2015). The aim is to keep the pituitary dose Dmax <50 Gy, but never at the expense of PTV coverage, which should always take priority.
For patients treated with radiotherapy alone, short course radiotherapy, with doses up to 40 Gy in 15 fractions over 3 weeks, has comparable survival rates to 60 Gy (Roa et al. 2004). Similar survival rates are also seen with 10 × 3.5 Gy fractions. Hypofractionated techniques are used in patients with poor prognostic features, such as advanced age and poor performance status (Weller et al. 2014).
Both 3-D conformal and virtual simulation radiotherapy techniques should be used depending on location, extent of disease, and prognosis.
MRI is the preferred imaging modality to assess response to radiotherapy usually 4–12 weeks after completion of treatment. Areas of increased contrast enhancement after treatment may be due to a reactive inflammatory process, rather than true disease progression (pseudo-progression) (Brandsma et al. 2008). Early radiological progression at the end of radiotherapy raises the probability of pseudo-progression (reported incidence of >60%); therefore, chemotherapy should be continued if clinically feasible and repeat imaging arranged after 8–12 weeks. If the follow-up scan does not demonstrate any further progression compared to the postradiotherapy baseline scan, then the initial tumor increase is termed pseudo progression and patients will continue with adjuvant chemotherapy as initially planned.
Response to chemotherapy is evaluated according to the Response Assessment in Neuro-Oncology (RANO) criteria (Wen et al. 2010). Tumor extent on T2 and FLAIR MRI sequences are evaluated along with the degree of contrast enhancement. Clinically, neurological function and steroid use is also included in the response assessment.
The increasing use of immunotherapy for relapsed HGG has led to the development of the immunotherapy response assessment for neuro-oncology (iRANO) criteria. This guidance incorporates the previously stated RANO criteria, with additional consideration for timing of immunotherapy commencement in order to accurately define progression at 3 and 6 months from treatment initiation (Okada et al. 2015).
MR spectroscopy and methionine PET imaging may help differentiate further between tumor recurrence and treatment-related changes but is not easily available in many centers (Hottinger et al. 2013).
Contrast-delayed MRI for calculating high resolution treatment response assessment maps (TRAMS) has been developed to differentiate between tumor recurrence and nontumor tissue (such as radiation necrosis) but require prospective validation (Zach et al. 2014). Differing rates of clearance and accumulation of contrast allow for discrimination between brain regions of high and low vascular activity.
MRI should be used to monitor the response to treatment, or as surveillance following completion of treatment. Radiological assessment at 3-monthly intervals is recommended for most patients with HGG in the initial period following therapy. Interval of assessment can be lengthened following a period of proven disease control.
The EORTC 22981 trial noted the recurrence rate for GBM and AO as 70–90% at 1 year and 89–98% at 2 years (Stupp et al. 2005). Care for patients with progressive disease should be tailored to individual performance status, age, previous treatment, and pattern of recurrence.
Suitable patients presenting with recurrent disease should be re-discussed with the neuroscience MDT to ascertain whether disease can safely be resected. While some studies have shown superior outcomes for patients who undergo repeat resection in comparison to historical controls, there is no randomized prospective data to support this; therefore, re-resection at the time of first or second progression remains a source of controversy (Helseth et al. 2010). Some analyses for prognostic indicators for survival do not identify re-resection at progression as a positive factor while other do (Ringel 2015; Gorlia et al. 2012). Surgery should, therefore, mainly be considered for good prognosis patients, with well-circumscribed lesions causing mass effect (Weller et al. 2014). Ideally, to allow for a clinically meaningful benefit, the interval since initial surgery should be 6 months or more. Insertion of chemotherapy (BCNU)-impregnated polymers at repeat surgery may also prolong survival in selected patients (Brem et al. 1995).
Re-irradiation can be considered for small recurrent tumors. Historically, re-irradiation rates have been extremely low due to the presumed increased risk of brain injury with higher radiation doses. With modern techniques, the rate of normal brain radiation necrosis has been shown to be low (Mayer and Sminia 2008). In a meta-analysis of over 300 patients, local recurrence were treated with a reported 6-month PFS of 28–39%, 1-year OS of 18–48%, and reduction in steroid use in 20–60% of cases (Nieder et al. 2008). Either LINAC-based stereotactic fractionated regimens (30 to 35 Gy) or stereotactic radiosurgery (15 to 20 Gy) can be considered (Romanelli et al. 2009). The use of particle therapies in the recurrent setting, e.g., heavy carbon ion radiotherapy, remains a further experimental approach, which may hold promise for the future.
Patients with a good performance status should be considered for second line chemotherapy as this can improve symptoms and quality of life. GBM patients and AA or AO who have previously been treated with TMZ/RT may be treated with nitrosurea-based single agent chemotherapy, PCV, or a temozolomide re-challenge (Perry et al. 2008).
High grade gliomas are highly vascular tumors. Targeting angiogenesis is perceived as a potential mode of controlling disease progression. Studies have revealed high response rates and a steroid-sparing effect with the use of Bevacizumab (+/− Irinotecan). PFS rates at 6 months vary in reported studies, ranging from 35% for GB to 56% for Grade III tumors (Bokstein et al. 2008; Desjardins et al. 2008).
In addition, inclusion in a locally or regionally open clinical phase I/II study should always be considered given the lack of a clearly defined optimal second line treatment and the extremely poor prognosis for patients with HGG at recurrence.
The diagnosis of gliomatosis cerebri, a term no longer used in the latest WHO classification of 2016, remains a clinically well-defined entity in current medical circles and is predominantly a radiologically descriptive diagnosis requiring histological verification. Scans should demonstrate diffuse hyperintensity on T2/FLAIR affecting at least three cerebral lobes. Typically, both cerebral hemispheres are affected, and disease can extend into the brainstem and basal ganglia. Histology for the majority of patients with gliomatosis cerebri is that of diffuse astrocytoma.
The prognosis for this subset of HGG is variable. Tumors with IDH1/2 mutations are associated with improved survival and the disease responds to both radiotherapy and chemotherapy (Herrlinger et al. 2002).
The role of radiotherapy as first line treatment is limited, due to the large target volumes required; hence, primary chemotherapy with Temozolomide or PCV has been reported as a common initial therapeutic approach (Sanson et al. 2004).
High grade glioma is associated with a poor prognosis in the elderly. Historical trials have excluded patients over the age of 65–75, complicating the development of clinical guidelines for this age group. Although the original Stupp trial did not show a benefit for adding TMZ to radiotherapy for patients over 70, a number of subsequent studies have illustrated an improvement in survival with combination therapy.
A recent meta-analysis comparing TMZ/RT with RT alone for patients over the age of 65 revealed a survival benefit for selected older patients, those with MGMT methylation status and good performance status and those who had undergone extensive surgical resection (Yin et al. 2013). A modest increase in reported toxicity was deemed acceptable.
A Nordic study assessing patients over the age of 65 compared standard radiotherapy (60 Gy in 1.8-2Gy fractions) with short course RT or TMZ. Survival rates for short course radiotherapy or single-agent TMZ were similar; however, standard radiotherapy (60 Gy) was associated with poorer outcome (Malmström et al. 2012).
A study assessing combined short course RT (40 Gy/15 fractions) with concomitant and adjuvant TMZ demonstrated a survival benefit for patients with unmethylated as well as methylated MGMT status (Perry et al. 2017). Doses of TMZ were the same as those recommended for younger cohorts.
Understanding molecular pathogenesis of malignant gliomas has led to an increased interest in targeted molecular therapies. EGFR overexpression and EGFR VIII mutation are associated with GB carrying the poorest prognosis. Targeting of these growth factors has been under investigation but has not yet demonstrated a survival benefit in the first line setting (Brandes et al. 2008). To date, gefitinib and erlotinib have also been studied with conflicting results.
In the setting of recurrence, rindopepimut is a promising peptide vaccine for GB patients that has been shown to improve survival in multicenter phase 2 studies (Reardon et al. 2014). Yet the recently published Phase 3 data of rindopepimut in the first line setting (ACT IV) did not demonstrate the anticipated survival advantage (Weller et al. 2017).
Activating BRAF mutations, occurring in up to 5% of GBs, are another potential therapeutic target and the subject of on-going research. The efficacy of dabrafenib (BRAF inhibitor) in combination with trametinib (MEK inhibitor) is being studied in HGG, along with other malignancies, in an international phase 2 trial (Subbiah et al. 2016).
Increasing interest has recently been paid to the use of immunotherapies due to the significant activity seen in malignancies such as melanoma and lung cancer. While not associated with currently available immune modulating therapies, HGG may potentially benefit from immunotherapeutic agents through the use of dendritic cell therapies, antiviral approaches, and checkpoint inhibitors. Several clinical trials exploring these targets are currently recruiting.
The benefit of Proton Beam radiotherapy for treating low and high grade gliomas has been explored but remains experimental in this setting. The use of Proton beam therapies in primary CNS tumors has the potential to allow for improved sparing of OARs and can reduce the radiotherapy-associated second malignancy rates, assuming the underlying prognosis of a patient justifies this given the limited accessibility and high associated costs to date.
The optimal management of high grade glioma remains a clinical challenge. Maximal safe surgical resection followed by chemoradiation remains the internationally accepted standard of care in patients with good performance status for newly diagnosed HGG. However, overall prognosis remains poor, specifically in WHO grade IV tumors. As the likelihood of presenting with HGG increases with advancing age, special considerations are needed in order to optimally manage the cohort of elderly patients with preexisting multiple comorbidities or cognitive impairment without causing unacceptable toxicity.
Despite primary intensive multimodality treatment, recurrence rates remain extremely high in HGG indicating an urgent need for further studies to improve outcomes. The updated histological classification has allowed for more accurate prognostication between different disease subtypes. In the era of molecular/DNA analyses of tumors, targeted and immunotherapeutic agents may provide an exciting target for future treatments and allow for a personalized approach to the management. Along with targeting oncogenic proteins, dendritic cell therapies, antiviral approaches, and checkpoint inhibitors, a substantial number of early phase clinical studies are currently open to recruitment moving towards the concept of “personalized medicine” in HGG. These need to be supported by the entire neuro-oncology community if any progress in this challenging area can be made in the future (Theeler and Gilbert 2015).
- Armstrong GT, Jain N, Liu W, Merchant TE, Stovall M, Srivastava DK, Gurney JG, Packer RJ, Robison LL, Krull KR. Region-specific radiotherapy and neuropsychological outcomes in adult survivors of childhood CNS malignancies. Neuro-Oncology. 2010;12(11):1173–86.CrossRefPubMedPubMedCentralGoogle Scholar
- Avril T, Saikali S, Vauléon E, Jary A, Hamlat A, De Tayrac M, Mosser J, Quillien V. Distinct effects of human glioblastoma immunoregulatory molecules programmed cell death ligand-1 (PDL-1) and indoleamine 2, 3-dioxygenase (IDO) on tumour-specific T cell functions. J Neuroimmunol. 2010;225(1):22–33.CrossRefPubMedGoogle Scholar
- Brem H, Piantadosi S, Burger PC, Walker M, Selker R, Vick NA, Black K, Sisti M, Brem S, Mohr G, Muller P. Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. Lancet. 1995;345(8956):1008–12.CrossRefPubMedGoogle Scholar
- Chemaitilly W, Li Z, Huang S, Ness KK, Clark KL, Green DM, Barnes N, Armstrong GT, Krasin MJ, Srivastava DK, Pui CH. Anterior hypopituitarism in adult survivors of childhood cancers treated with cranial radiotherapy: a report from the St Jude lifetime cohort study. J Clin Oncol. 2015;33(5):492–500.CrossRefPubMedPubMedCentralGoogle Scholar
- Dixit S, Hingorani M, Achawal S, Scott I. Retrospective comparison of chemoradiotherapy followed by adjuvant chemotherapy, with or without previous gliadel implantation (carmustine) after initial surgery in patients with newly diagnosed high-grade gliomas: in regard to noel et al. Int J Radiat Oncol Biol Phys. 2011;81(5):1593.CrossRefPubMedGoogle Scholar
- Gondi V, Tolakanahalli R, Mehta MP, Tewatia D, Rowley H, Kuo JS, Khuntia D, Tomé WA. Hippocampal-sparing whole-brain radiotherapy: a “how-to” technique using helical tomotherapy and linear accelerator–based intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys. 2010;78(4):1244–52.CrossRefPubMedPubMedCentralGoogle Scholar
- Hartmann C, Meyer J, Balss J, Capper D, Mueller W, Christians A, Felsberg J, Wolter M, Mawrin C, Wick W, Weller M. Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial differentiation and age: a study of 1,010 diffuse gliomas. Acta Neuropathol. 2009;118(4):469–74.CrossRefPubMedGoogle Scholar
- Jenkins RB, Blair H, Ballman KV, Giannini C, Arusell RM, Law M, Flynn H, Passe S, Felten S, Brown PD, Shaw EG. A t (1; 19)(q10; p10) mediates the combined deletions of 1p and 19q and predicts a better prognosis of patients with oligodendroglioma. Cancer Res. 2006;66(20):9852–61.CrossRefPubMedGoogle Scholar
- Malmström A, Grønberg BH, Marosi C, Stupp R, Frappaz D, Schultz H, Abacioglu U, Tavelin B, Lhermitte B, Hegi ME, Rosell J. Temozolomide versus standard 6-week radiotherapy versus hypofractionated radiotherapy in patients older than 60 years with glioblastoma: the Nordic randomised, phase 3 trial. Lancet Oncol. 2012;13(9):916–26.CrossRefPubMedGoogle Scholar
- Paulsson AK, McMullen KP, Peiffer AM, Hinson WH, Kearns WT, Johnson AJ, Lesser GJ, Ellis TL, Tatter SB, Debinski W, Shaw EG. Limited margins using modern radiotherapy techniques does not increase marginal failure rate of glioblastoma. Am J Clin Oncol. 2014;37(2):177.CrossRefPubMedPubMedCentralGoogle Scholar
- Reardon DA, Schuster J, Tran DD, Fink KL, Nabors LB, Li G, Bota DA, Lukas RV, Desjardins A, Ashby LS, Duic JP. ReACT: overall survival from a randomized phase II study of rindopepimut (CDX-110) plus bevacizumab in relapsed glioblastoma. Neuro Oncol. 2014;16(Suppl 5):v116Google Scholar
- Ringel F. SN1 study group, Pape H, SN1 study group, Sabel M, SN1 study group, Krex D, SN1 study group, bock HC, SN1 study group, Misch M. Clinical benefit from resection of recurrent glioblastomas: results of a multicenter study including 503 patients with recurrent glioblastomas undergoing surgical resection. Neuro-Oncology. 2015;18(1):96–104.CrossRefPubMedPubMedCentralGoogle Scholar
- Romanelli P, Conti A, Pontoriero A, Ricciardi GK, Tomasello F, De Renzis C, Innocenzi G, Esposito V, Cantore G. Role of stereotactic radiosurgery and fractionated stereotactic radiotherapy for the treatment of recurrent glioblastoma multiforme. Neurosurg Focus. 2009;27(6):E8.CrossRefPubMedGoogle Scholar
- Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ, Janzer RC, Ludwin SK, Allgeier A, Fisher B, Belanger K, Hau P. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol. 2009;10(5):459–66.CrossRefGoogle Scholar
- Stupp R, Hegi ME, Gorlia T, Erridge SC, Perry J, Hong YK, Aldape KD, Lhermitte B, Pietsch T, Grujicic D, Steinbach JP. Cilengitide combined with standard treatment for patients with newly diagnosed glioblastoma with methylated MGMT promoter (CENTRIC EORTC 26071-22072 study): a multicentre, randomised, open-label, phase 3 trial. Lancet Oncol. 2014;15(10):1100–8.CrossRefPubMedGoogle Scholar
- Subbiah V, Bang YJ, Lassen UN, Wainberg ZA, Soria JC, Wen PY, Zenz T, Moreau P, Brunsvig P, De Braud FG, De Greve J. ROAR: a phase 2, open-label study in patients (pts) with BRAF V600E–mutated rare cancers to investigate the efficacy and safety of dabrafenib (D) and trametinib (T) combination therapy. J Clin Oncol. 2016;34(15).Google Scholar
- Van Den Bent MJ, Erridge S, Vogelbaum MA, Nowak AK, Sanson M, Brandes AA, Wick W, Clement PM, Baurain JF, Mason WP, Wheeler H. Results of the interim analysis of the EORTC randomized phase III CATNON trial on concurrent and adjuvant temozolomide in anaplastic glioma without 1p/19q co-deletion: an Intergroup trial. Lancet. 2017;390:1645–1653Google Scholar
- Weller M, Butowski N, Tran DD, Recht LD, Lim M, Hirte H, Ashby L, Mechtler L, Goldlust SA, Iwamoto F, Drappatz J. Rindopepimut with temozolomide for patients with newly diagnosed, EGFRvIII-expressing glioblastoma (ACT IV): a randomised, double-blind, international phase 3 trial. Lancet Oncol. 2017;18(10):1373–85.CrossRefPubMedGoogle Scholar
- Wick W, Hartmann C, Engel C, Stoffels M, Felsberg J, Stockhammer F, Sabel MC, Koeppen S, Ketter R, Meyermann R, Rapp M. NOA-04 randomized phase III trial of sequential radiochemotherapy of anaplastic glioma with procarbazine, lomustine, and vincristine or temozolomide. J Clin Oncol. 2009;27(35):5874–80.CrossRefPubMedGoogle Scholar
- Zhao F, Li M, Kong L, Zhang G, Yu J. Delineation of radiation therapy target volumes for patients with postoperative glioblastoma: a review. OncoTargets Ther. 2016;9:3197–204.Google Scholar