Diffuse intrinsic pontine glioma

INTRODUCTION — Brainstem gliomas are characterized by heterogeneous biologic behavior, ranging from low-grade tumors needing little treatment to those that are rapidly fatal despite aggressive therapy [1,2]. Prognosis and treatment depend upon both the clinical symptoms and their duration, the location of the tumor within the brainstem, and, increasingly, the mutational profile.

Approximately 80 percent of pediatric brainstem gliomas arise within the pons, while the remaining 20 percent arise in the medulla, midbrain, or cervicomedullary junction (figure 1) [3-8]. The majority of pontine tumors are diffuse intrinsic brainstem gliomas, which are usually high grade, locally infiltrative, and have a uniformly poor prognosis [9]. Histologically, these tumors are usually World Health Organization (WHO) grade 3 (anaplastic) astrocytomas or glioblastoma (WHO grade 4). However, patients with WHO grade 2 tumors identified by biopsy do not have an improved prognosis. (See "Classification and pathologic diagnosis of gliomas, glioneuronal tumors, and neuronal tumors".)

By contrast, most nonpontine tumors involving the cervicomedullary junction and tectum, as well as focal, cystic, and dorsal exophytic lesions, are low-grade astrocytomas, mostly grade 1 pilocytic astrocytomas [6]. These are discrete, well-circumscribed tumors, often without evidence of locally invasive growth or edema [10]. Approximately 10 to 20 percent of nonpontine gliomas will be high grade and are treated similarly to diffuse intrinsic pontine gliomas.

Diffuse intrinsic pontine gliomas will be reviewed here. Gliomas arising from other sites within the brainstem are discussed separately. (See "Focal brainstem glioma".)

EPIDEMIOLOGY — Gliomas arising in the brainstem (midbrain, pons, and medulla oblongata) account for 10 to 20 percent of all central nervous system tumors in children and approximately one-third of high-grade gliomas in children [11]. Brainstem gliomas are more common in children than adults [3,4,12,13]. In the United States, for example, there are approximately 300 pediatric cases and 100 adult cases reported each year. In children, the median age at diagnosis is five to nine years, and there is a slight female sex predominance [11].

PATHOLOGY

Histopathology and grade — The most commonly used classification system for central nervous system tumors is that of the World Health Organization (WHO) [14]. Astrocytomas account for greater than 95 percent of all brainstem lesions. Most centers classify these tumors based on the WHO criteria, which encourage use of integrated and layered diagnoses to accommodate histologic and genetic parameters into a single diagnosis (algorithm 1 and table 1) (see "Classification and pathologic diagnosis of gliomas, glioneuronal tumors, and neuronal tumors", section on 'Histopathologic and molecular classification'). For diffuse intrinsic pontine tumors, clinical and radiographic characteristics continue to be the most important predictors of outcome.

When biopsied, diffuse intrinsic pontine gliomas are usually high-grade astrocytomas, although up to one-quarter appear low grade on classic histologic features, and nearly all tumors progress rapidly. Importantly, histopathologic grade does not correlate with prognosis in diffuse intrinsic pontine gliomas, and even low-grade diffuse pontine lesions behave aggressively and carry a similarly poor prognosis compared with high-grade tumors.

Glial tumors of the pons have rarely metastasized to distant sites at the time of diagnosis. Instead, they tend to progress along established fiber tracts, especially into the thalamus and cerebellum. Spinal seeding is rare but can be observed in the end stages of the disease [15,16]. With an average life expectancy of less than a year, the absence of metastatic disease may be the result of the rapidly fatal course of the primary tumor, rather than a lack of the metastatic potential of these tumors.

Tumors of the cervicomedullary junction, as well as focal and dorsal exophytic lesions, tend to be low grade based upon the WHO classification. When a high-grade component of one of these lesions is observed, the tumor tends to behave similarly to high-grade glial lesions in other parts of the central nervous system.

Molecular pathogenesis — Significant insights have been gained into the molecular biology of diffuse intrinsic pontine gliomas using biopsy tissue and autopsy material [17,18]. Despite similar histopathology, adult and pediatric diffuse gliomas are now recognized to contain distinct underlying genetic events.

●Mutually exclusive mutations in either H3F3A, one of two genes encoding the histone H3.3 variant, or HIST3H1B, one of several genes encoding histone H3.1, have been identified in nearly 80 percent of diffuse intrinsic pontine gliomas and appear to be present in all tumor cells [19-21]. Both mutations involve a substitution of lysine 27 to methionine (H3 K27M), which results in altered binding of mutant histone H3 to polycomb repressive complex 2 (PRC2), a key developmental regulator of gene expression [22]. H3F3A-mutant tumors may also harbor mutations in protein phosphatase, magnesium/manganese-dependent (PPM1D), which alter the DNA damage response pathway [23].

●Amplification and overexpression of the platelet-derived growth factor receptor (PDGFR) A and c-Met proteins seem to be important molecular events in the transformation of cells to a malignant phenotype [24], seen more commonly in H3.3-mutant tumors. Activating mutations in phosphatidylinositol 3-kinase (PI3K) are identified in approximately 15 percent of diffuse intrinsic pontine gliomas [25] and seem to be more common in H3.1-mutant tumors.

●Recurrent somatic mutations in activin A receptor, type 1 (ACVR1), a gene that encodes a bone morphogenetic protein (BMP) type 1 receptor, have been identified in up to one-third of pediatric diffuse intrinsic pontine gliomas [26-29]. Mutations result in activation of the BMP pathway and increased expression of a variety of downstream effectors, including Smad proteins and members of the ID gene family, which may represent potential therapeutic targets. These mutations are often associated with concurrent mutations in PI3K [25,28] and the H3.1 gene [28,29] and may be associated with slightly better prognosis [30].

●Mutations in isocitrate dehydrogenase type 1 (IDH1) or type 2 (IDH2), which define a large proportion of adult grade 2 and 3 diffuse gliomas, are extremely rare in pediatric diffuse intrinsic pontine gliomas [31,32]. By contrast, IDH1/2 mutations, mostly non-R132H variants, may be present in up to one-quarter of adult brainstem gliomas, are associated with improved prognosis [32,33], and may benefit from adjuvant chemotherapy in addition to radiation therapy [34]. (See "Classification and pathologic diagnosis of gliomas, glioneuronal tumors, and neuronal tumors", section on 'IDH1/IDH2 mutation'.)

Relevant mouse models based on these molecular insights have been developed to aid in the identification of targeted therapies for diffuse intrinsic pontine gliomas. (See 'Investigational therapies' below.)

Diffuse midline glioma, H3 K27-altered — Diffuse midline glioma, H3 K27-altered was a new diagnostic entity as of the 2016 revised version of the WHO classification of brain tumors [35]. As discussed above, approximately 80 percent of biopsied diffuse intrinsic pontine gliomas harbor mutations at position K27 in one of several histone-encoding genes. (See 'Molecular pathogenesis' above.)

Per the WHO classification, H3 K27-altered midline gliomas are high-grade gliomas, by definition WHO grade 4, associated with local infiltration and a poor prognosis, even when they have the histologic appearance of a low-grade glioma [14,36]. Although the pons is the most common location, H3 K27-altered gliomas also occur in the thalamus, spinal cord, and other midline sites and can affect both children and adults [33,37,38]. The natural history of these tumors in locations outside the brainstem and in adults is not yet well defined and may be more heterogeneous than that of H3 K27-altered diffuse intrinsic pontine gliomas [39,40].

Importantly, H3 K27-altered diffuse midline glioma is a pathologic diagnosis and is not synonymous with the clinical and radiologic entity of diffuse intrinsic pontine glioma, which does not require biopsy for diagnosis (and is often not biopsied) and behaves uniformly poorly, regardless of histologic grade or H3 K27M status.

Advanced sequencing techniques on circulating tumor cells in blood and cerebrospinal fluid may eventually aid in noninvasive identification of H3 K27-altered gliomas [41].

CLINICAL PRESENTATION — Diffuse intrinsic pontine gliomas can present with varied symptoms depending on the location of the lesion. These include:

●Cranial nerve palsies, long tract signs (eg, hemiparesis), and ataxia in over 50 percent of patients. Cranial nerves VI and VII are most commonly affected, but III, IV, IX, and X may also be involved.

●Hydrocephalus with elevated intracranial pressure (ICP) is observed in less than 10 percent of patients at presentation.

●Significant intratumoral hemorrhage can be present in approximately 6 percent of patients at the time of diagnosis, although small punctate hemorrhages can be seen in a high percentage of cases [42]. Symptomatic hemorrhage may eventually occur in up to 20 percent of children, usually in necrotic areas of tumor.

None of these symptoms are pathognomonic of diffuse intrinsic pontine gliomas. Other brainstem tumors can present with similar symptoms. (See "Focal brainstem glioma".)

●Cervicomedullary tumors typically present with lower cranial nerve dysfunction and ataxia and have a prolonged clinical course, often progressing over years rather than days or weeks.

●Focal tumors tend to interfere with the cranial nerves adjacent to the site of origin.

●Unlike other brainstem tumors, tectal plate lesions present with hydrocephalus in over 90 percent of cases. Although these tumors are rarely biopsied, those that have been studied are almost always low-grade astrocytomas.

In contrast to focal brainstem gliomas, diffuse intrinsic pontine gliomas usually present with a short duration of symptoms (typically less than three months). The duration of symptoms in the context of classic radiographic findings is an important predictor of prognosis in brainstem tumors. Patients with symptoms of greater than six months' duration at the time of diagnosis tend to do much better than patients with a shorter duration. (See "Focal brainstem glioma", section on 'Clinical features'.)

NEUROIMAGING — Imaging studies, particularly magnetic resonance imaging (MRI) and, less frequently, computed tomography (CT), are the standard method of diagnosis and classification for brainstem tumors [6]. The brainstem is one of the most eloquent areas of the brain, making biopsy difficult in centers without expertise in this area. (See 'Surgery' below.)

Diffuse intrinsic pontine gliomas have a typical appearance on CT scans with hypodense to isodense appearance and variable contrast enhancement. Calcium is rarely identified within the tumor. MRI is superior to CT for defining these lesions. On MRI, diffuse pontine lesions are expansile, typically hypointense on T1- and hyperintense on T2-weighted images (image 1).

High-grade tumors typically invade adjacent areas including the medulla or midbrain. Axial growth is commonly observed through the cerebellar peduncles and into the cerebellum, and the tumors expand the pons, instead of displacing it. They frequently have exophytic components into the prepontine cistern and often encircle the basilar artery. On sagittal images, the tumors appear to respect the pontomedullary boundary.

Contrast enhancement within the tumors can be variable and may be ring-enhancing with central areas of necrosis. The apparent well-circumscribed appearance of ring-enhancing lesions on MRI is often not confirmed at the time of surgery or autopsy and does not indicate a lower-grade lesion or more benign clinical course. Many diffuse pontine tumors have no contrast enhancement. Tumors that demonstrate diffuse uniform enhancement are more likely low grade, are usually associated with prolonged symptoms, and may have a better outcome.

Magnetic resonance spectroscopy (MRS) with estimation of choline-to-N-acetylaspartate (Cho:NAA) ratios is another noninvasive method that can be used to support the diagnosis and may provide prognostic information [43,44]. In one series, children with a maximum Cho:NAA ratio greater than 4.5 had a median survival of 22 weeks, and all 13 patients died by 63 weeks. By contrast, the projected survival of patients with a Cho:NAA ratio ≤4.5 was significantly better, 50 percent at 63 weeks [45].

Other potentially useful techniques to further assist in imaging-based diagnoses for brainstem tumors include rapid diffusion MRI, thallium single-photon emission computed tomography (SPECT) [46], and positron emission tomography (PET) [47]. As an example, 111In-pentetreotide binds the somatostatin receptor, which is frequently expressed in gliomas. Scanning with this agent can both help support the diagnosis and, if labeled, potentially target radiation therapy to these lesions [48]. While these methods continue to undergo further development and refinement, they are being used more frequently to assist MRI-based diagnosis, particularly in difficult or unclear cases.

While MRI has become the standard method of evaluation of diffuse intrinsic pontine gliomas, the heterogeneous signal characteristics of these lesions and interobserver variability make serial assessment difficult [49]. Evaluation of new therapies should therefore be based upon overall survival rather than MRI changes whenever possible [50].

DIAGNOSIS — Advances in imaging techniques and the difficulty in obtaining an adequate biopsy sample make the interpretation of the imaging studies critical in the differential diagnosis. If imaging findings are typical, biopsy is not usually necessary to confirm the diagnosis and should only be performed in the context of a formal clinical trial [51].

Patients with a long duration of symptoms and atypical clinical or imaging findings have an increased likelihood of having a lesion other than diffuse intrinsic pontine glioma. Such patients, as well as adolescents and adults, should be considered for biopsy, given the possibility of an IDH-mutant glioma. (See 'Surgery' below.)

Differential diagnosis — Vascular malformations, encephalitis, rare parasitic cysts, demyelinating disorders (eg, multiple sclerosis), and hamartomas in patients with neurofibromatosis are the most common causes of nonneoplastic lesions in the brainstem. Rare examples of metastases from extracranial carcinomas in adults have also been reported in the pons [52].

TREATMENT — The management of a patient with a diffuse intrinsic pontine glioma includes control of peritumoral edema, as well as specific measures directed at the tumor (surgery, radiation therapy, and chemotherapy). Participation in clinical trials is encouraged, particularly as the molecular pathogenesis of these tumors is unravelled and targeted therapies begin to enter clinical trials. (See 'Molecular pathogenesis' above and 'Investigational therapies' below.)

Glucocorticoids — The administration of glucocorticoids (eg, dexamethasone) is an important initial step in the treatment of patients with high-grade brainstem tumors. Peritumoral edema often contributes prominently to symptoms, which can improve rapidly after steroid therapy. Although steroids do not treat the underlying problem, improved clinical performance can be an important matter for quality of life in patients with a limited life span.

Unfortunately, many patients require prolonged administration of dexamethasone, resulting in significant steroid-related complications. In some cases, bevacizumab, a monoclonal antibody against vascular endothelial growth factor (VEGF), has been used to modulate peritumoral edema without the complications of glucocorticoids [53]. (See "Management of vasogenic edema in patients with primary and metastatic brain tumors", section on 'Refractory edema'.)

Surgery — Unless the clinical or imaging picture is atypical, surgery is usually not recommended in children with a clinical diagnosis of diffuse brainstem glioma outside of formal clinical trials [54]. The morbidity associated with surgery in this eloquent region of the brain and the sampling error associated with biopsies preclude routine biopsy to identify patients with lesions that are not high-grade astrocytomas [10,55,56]. Pontine biopsy should only be performed by surgeons with expertise in this procedure and ideally in the context of a clinical trial where the analysis will influence therapy [51].

Adults are more likely to benefit from a biopsy. In this setting, determining the tissue diagnosis may improve prognostication and direct therapy, such as in the case of IDH-mutant glioma. In addition, up to 30 percent of biopsied patients had a diagnosis other than astrocytoma (eg, lymphoma, ependymoma, or infection) in studies that were completed both before and after availability of MRI [57,58].

Given the hazards of the surgical approach and rarity of brainstem lesions, referral to a highly specialized and dedicated neurosurgical team should be made before treatment is initiated, particularly for lesions that lack some of the classic clinical and/or imaging features [59].

Advances in neurosurgical techniques have permitted biopsy of diffuse intrinsic pontine gliomas to be performed safely [51,60-62]. Pretherapy biopsy is being performed in the context of formal clinical trials to identify molecular pathways aberrant in these tumors and to use this information to guide treatment [51]. The feasibility and safety of biopsy has been demonstrated in a multicenter prospective study of 53 children with presumed diffuse intrinsic pontine glioma, in which 50 patients underwent stereotactic biopsy [63]. Two patients had procedural grade 3 adverse events (hypertension, apnea), and one patient developed a hemiparesis with incomplete recovery. There were no biopsy-related deaths, and all but one patient went on to receive planned radiation therapy at a median of 10 days postbiopsy. All but four samples yielded sufficient tissue for molecular analysis including O6-methylguanine-DNA methyltransferase (MGMT) methylation and whole genome sequencing.

Radiation therapy — Radiation therapy is the only treatment that appears to alter the clinical course of diffuse intrinsic pontine gliomas. Despite multiple trials of dose escalation, altered fractionation, and radiosensitization, none of these modulations have been proven more effective than conventionally delivered radiation therapy.

Radiation therapy treatment fields are typically restricted to the tumor volume plus 1 to 2 cm of adjacent brainstem tissue. The standard treatment dose is 1.8 Gy daily given five days per week, to a total dose of between 54 and 59.4 Gy. Patients with severe symptoms may require urgent initiation of radiation therapy.

Although the degree of tumor shrinkage induced by radiation therapy can be dramatic, the response is generally transient, and patients are not cured by this approach. The median survival is approximately 10 months, and the two-year overall survival rate is less than 10 percent.

In older studies, the duration of response was correlated with the tumor grade, as well as the bulk of disease, although this has not been demonstrated in more recent studies [64]. The time to progression may be prolonged in patients with neurofibromatosis type 1 and in children as compared with older adults. Treatment failures are usually local and occur within the radiation therapy field. Although up to one-third of patients have evidence of central nervous system dissemination at the time of progression, most symptoms are referred to the primary site [65,66]. (See "Neurofibromatosis type 1 (NF1): Pathogenesis, clinical features, and diagnosis".)

Due to the poor prognosis in these patients, hypofractionated radiation therapy has been evaluated to minimize the time spent in treatment [67-71]. In the largest of the trials among 253 children with newly diagnosed diffuse intrinsic pontine glioma, overall survival ranged from 8.2 to 9.6 months and was similar across three randomized treatment groups: 39 Gy in 13 fractions, 45 Gy in 15 fractions, and 54 Gy in 30 fractions [71]. Acute and delayed side effects and toxicities were also similar across groups.

Alternative radiation therapy techniques — A number of alternative approaches have been used to try to improve the effectiveness of radiation therapy.

●Dose escalation, using multiple daily treatments with smaller than conventional fraction sizes, appeared to increase the radiographic response (table 2). However, no overall survival advantage was observed, and many patients experienced significant toxicity due to prolonged steroid dependence [72-74].

●The addition of radiation sensitizers or concomitant or preradiation chemotherapy generally has failed to improve the time to disease progression [75-78].

●Brachytherapy has been used to increase the delivery of radiation to the tumor after conventional-dose radiation therapy. In one report, there was minimal morbidity related to placement of 125-I in nine patients with diffuse intrinsic pontine glioma; however, this approach did not improve the response rate or survival outcome [79].

●Stereotactic radiosurgery has been used in brainstem lesions, although great care is required because of the potential for radionecrosis in vital areas [80]. There are no data that this technique improves the outcome in this disease.

Fractionated reirradiation of pontine gliomas at the time of progression may be considered to temporarily extend palliation, although increased toxicities may result from this approach [81,82]. In a small prospective dose-escalation study, the lowest dose level (24 Gy in 12 fractions) was well tolerated and achieved imaging, clinical, and/or quality-of-life improvement in most patients; in all 12 patients, progression-free and overall survival from the start of reirradiation were 4.5 and 5.8 months, respectively [83].

Chemotherapy — Chemotherapy has been ineffective in the treatment of children with diffuse intrinsic pontine gliomas. Numerous treatment protocols, including single chemotherapy agents, multidrug combination regimens, and high-dose therapy with stem cell rescue, have all been tested in both adults and children without clear evidence of benefit [84-86].

Although temozolomide has become part of standard therapy for most adult patients with high-grade gliomas, testing in combination with radiation for newly diagnosed diffuse intrinsic pontine glioma has demonstrated no improvement in activity compared with radiation alone and is associated with increased risk for side effects and toxicity [87-89]. However, temozolomide should be considered for patients with isocitrate dehydrogenase (IDH)-mutant brainstem tumors given the demonstrated survival benefit in IDH-mutant 1p/19q non-codeleted anaplastic glioma [34].

Newer regimens have combined chemotherapy and biologic modifiers [90], but no regimen has demonstrated better activity than radiation therapy alone. Most of these efforts have been based upon results of studies in patients with high-grade gliomas arising elsewhere in brain.

Investigational therapies — A number of emerging preclinical insights have the potential for rapid translation into clinical trials in diffuse intrinsic pontine glioma [17,91-96]. Examples include pharmacologic inhibition of histone demethylation [91] and use of chimeric antigen receptor (CAR)-expressing T cells with specific targeting against either GD2, a disialoganglioside that is highly expressed on the surface of H3 K27M-mutant glioma cells [96], or B7-H3, an immune regulatory protein expressed on diffuse intrinsic pontine glioma [97].

The investigational oral compound ONC201 is under active investigation in H3 K27-altered midline gliomas. The mechanism of action is not well defined but may involve mitochondrial metabolism and the stress response pathway [98]. In a molecularly unselected phase II trial of ONC201, a near-complete radiographic response was observed in a patient with a recurrent H3 K27M-mutant thalamic glioma [99]. A subsequent series of 18 patients with H3 K27M-mutant gliomas (including eight children with diffuse intrinsic pontine glioma) treated on a compassionate basis included 14 patients with recurrent disease after radiation therapy [100]. Among these patients, median progression-free survival was 14 weeks, and three adults remained on treatment and progression free with a median follow-up of 50 weeks. Follow-up of four children treated with adjuvant ONC201 after radiation is ongoing. Several phase II trials of ONC201 have been initiated based on these findings [101].

Local therapies are also being studied, including DNX-2401, an oncolytic adenovirus administered by catheter infusion into the cerebellar peduncle [102].

Studies of systemic checkpoint inhibitor immunotherapies have thus far shown no clear signal of efficacy [103].

PROGNOSIS — The prognosis of diffuse intrinsic pontine glioma is poor, with a median overall survival of 10 to 11 months [86,104-106]. Five-year survival is less than 3 percent, and many long-term survivors have evidence of moderate or severe cognitive impairment, likely as a consequence of radiation therapy [106,107].

Several studies have suggested that younger patients (age <3 years) and older patients (age >10 years) may have a somewhat better prognosis [104,107-110]. Other favorable prognostic factors include long interval between the onset of symptoms and diagnosis, lack of pontine cranial nerve palsies, and atypical radiologic characteristics (eg, lack of enhancement) [84,104,106,107,111,112].

While both high- and low-grade diffuse pontine lesions do poorly in children, adults with low-grade histology may have a somewhat longer time to progression, although most eventually succumb to their disease [113,114].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Primary brain tumors".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology – Diffuse intrinsic pontine gliomas are rare tumors that are more frequent in young children than in adults, with a median age of onset of five to nine years. (See 'Epidemiology' above and 'Clinical presentation' above.)

●Molecular pathogenesis – Nearly 80 percent of biopsied diffuse intrinsic pontine gliomas harbor H3 K27M mutations in one of several histone-encoding genes. Other important molecular changes in subsets of tumors include alterations in the platelet-derived growth factor receptor (PDGFR), PI3K mutations, and recurrent somatic mutations in ACVR1. (See 'Molecular pathogenesis' above and 'Diffuse midline glioma, H3 K27-altered' above.)

●Histopathology and grade – Histopathologic grade does not correlate with prognosis in diffuse intrinsic pontine gliomas, and even low-grade diffuse intrinsic pontine lesions behave aggressively and carry a similarly poor prognosis compared with high-grade tumors. (See 'Histopathology and grade' above.)

●Diagnosis – Because of their location in an eloquent area of the brain, diagnosis is usually based upon imaging studies (image 1) and classic clinical features. Surgery or surgical biopsy is not routinely indicated, unless atypical features are present, IDH-mutant glioma is in the differential diagnosis, or patients are part of a formal clinical study. (See 'Neuroimaging' above and 'Surgery' above.)

●Treatment – The management of a patient with a diffuse intrinsic pontine glioma includes control of peritumoral edema and radiation therapy. Participation in clinical trials is encouraged.

•Glucocorticoids are used to control tumor-associated edema in symptomatic patients. (See "Management of vasogenic edema in patients with primary and metastatic brain tumors", section on 'Symptomatic treatment'.)

•We recommend external beam radiation therapy as the primary approach to antitumor therapy (Grade 1B). (See 'Radiation therapy' above.)

Our typical approach is to give a total dose of 54 to 59.4 Gy, in fractions of 1.8 Gy five times per week. Alternative radiation therapy techniques (hyperfractionation, radiation sensitizers, brachytherapy) have not been demonstrated to improve outcomes.

•Chemotherapy does not have an established role in the management of patients with diffuse intrinsic pontine glioma. Currently, chemotherapy in conjunction with radiation therapy is considered investigational and should be provided in the context of a clinical trial.

We recommend against use of temozolomide in patients with diffuse intrinsic pontine gliomas (Grade 1C). Two phase II multicenter trials have found no benefit compared with historical controls, and the risk of side effects and toxicity is increased compared with radiation alone. This recommendation does not apply to the minority of brainstem gliomas that are isocitrate dehydrogenase (IDH) mutant, for which temozolomide may be considered. (See 'Chemotherapy' above.)

●Prognosis – Despite aggressive therapy, the prognosis remains poor, with an average survival of less than one year. (See 'Prognosis' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Mark Kieran, MD, PhD, who contributed to an earlier version of this topic review.