Clinical manifestations

Presentation and course

Diffuse intrinsic brainstem gliomas classically present with the triad of cranial neuropathies, ataxia, and long tract signs (66). The type of neurologic compromise present is obviously dependent on the location of the tumor. Because the majority of diffuse intrinsic brainstem gliomas arise in the pons, sixth and seventh nerve dysfunction are most frequent. Lower cranial nerve deficits may also occur but are infrequently the first sign of diffuse intrinsic lesions. Rarely, patients will present with an isolated cranial neuropathy, such as peripheral seventh nerve palsy. The majority of patients will have some type of gait disturbance at the time of diagnosis. This disturbance is often an intermixture of long tract signs and ataxia. The classical presentation of crossed hemiparesis, with a seventh nerve palsy on one side and arm and leg weakness on the other side, clearly points to intrinsic lateralized brainstem involvement. Although the majority of tumors occur in young school-age children, tumors may arise in adolescents and young adults. Older children, comprising about 20% of all pediatric age patients with diffuse midline gliomas, tend to present with a longer symptom duration than younger children (21).

The duration between the onset of symptoms and the diagnosis of a brainstem lesion has changed with the improvement in neuroimaging techniques (06; 43). The majority of patients are now being diagnosed within two months of the insidious onset of symptoms, whereas prior to CT and MRI, most patients were diagnosed between three and six months of clinical presentation. Approximately one third of patients with brainstem tumors will have headache associated with nausea and vomiting. Although brainstem tumors can be bulky at diagnosis, with both anterior and posterior extension, only one third of patients will have hydrocephalus. Interestingly, patients may also experience significant personality changes prior to diagnosis. The reasons for such personality changes are unclear but may be due, in part, to the interruption of thalamic projections. In children less than five years of age, unexplained irritability is a frequent initial symptom.

The presentation of patients with localized cervicomedullary lesions differs from that of patients with diffuse intrinsic tumors. The patients with cervicomedullary lesions often have long histories of nonspecific headaches and vomiting (81). The headaches and vomiting rarely just occur early in the morning and are likely to occur at variable times during the day. Vomiting is usually projectile and is often unrelated to food intake. At time of diagnosis, patients may have little in the way of focal neurologic deficits or may have swallowing difficulties associated with lower cranial nerve (9, 10, 11, 12) dysfunctions. Swallowing difficulties and drooling tend to occur in young children.

Tumors of the midbrain or upper tectum often present with little, if any, focal neurologic deficits (85; 69; 75). Tectal masses most commonly result in hydrocephalus with associated headaches, nausea, and vomiting. After diversion of cerebrospinal fluid, there may be a complete disappearance of symptoms. Less frequently, patients with tectal lesions will have frank extraocular movement dysfunction including Parinaud syndrome (paralysis of upgaze, lid retraction, convergence nystagmus, and pupils that react better to accommodation than to light).

Occasionally, focal pontine lesions will arise, presenting with isolated seventh nerve palsies. These lesions are usually pilocytic astrocytomas and seem to carry a better prognosis (18).

Prognosis and complications

The overall prognosis for the majority of patients with brainstem gliomas is poor. Patients with diffuse intrinsic brainstem gliomas rarely survive as less than 10% of patients are alive and free of progressive disease 18 months following diagnosis and treatment with radiotherapy (51; 61; 62; 37). Most patients relapse 5 to 12 months after initial treatment. Up to 20% of children will have evidence of leptomeningeal dissemination prior to death (33). Younger children, especially those less than three years of age at diagnosis, may be more responsive to chemotherapy and have a better prognosis (11).

The prognoses for patients with more focal, often histologically benign, lesions are more favorable. After standard radiotherapy, nearly 50% of patients with focal lesions can be expected to be alive and free of progressive disease five years after diagnosis (01). Patients with cervicomedullary lesion tend to have a relatively favorable prognosis. In some series, there is short term disease control (three years) in the majority of patients (20). However, it is unclear whether this translates into long-term, disease-free survival. Similarly, children with isolated tectal lesions tend to have a better prognosis, with the majority of patients being alive and free of progressive disease for at least two to three years following initial diagnosis, many of whom have not received any specific anticancer treatment (85; 69; 75).

Clinical vignette

A 12-year-old boy developed double vision approximately two weeks prior to this evaluation. In addition, he had increased clumsiness, and, during the week prior to his assessment, had developed some difficulty swallowing. His speech had become garbled, and he had mild but constant headaches. Two days before the evaluation there had been some associated nausea and intermittent vomiting.

On examination, the child was alert and oriented. His pupils were equal and reactive, and his discs were flat. Extraocular movements were abnormal, and the child had a right sixth nerve palsy. Gaze to the right was limited, with incomplete abduction of the left eye. There was a mild right facial weakness. His gag was decreased, and the uvula pulled slightly to the left. Tongue movements were normal. On motor and coordination testing, dysmetria was more marked in the right upper extremity than on the left. When the child walked, he tended to fall to the right. There was mild drift of the left upper extremity. Reflexes were increased on the left as compared to the right.

The CT scan disclosed marked enlargement of the brainstem, which was hypodense as compared to normal brain. The fourth ventricle was flat and pushed posteriorly. There was no associated hydrocephalus. MRI scan showed a large, ill-defined hypointense abnormality on T1-weighted images that was hyperintense on T2-weighted images. There was minimal contrast enhancement.

Biological basis

Etiology and pathogenesis

The etiology for the vast majority of patients with brainstem gliomas remains unknown. There is a higher incidence of tumors in children with neurofibromatosis type 1. Usually these are low-grade gliomas, but anaplastic piloid tumors may occur with concomitant CDKN2A/B and ATRX mutations (74).

The biological underpinnings of brainstem gliomas are better understood. Pediatric brainstem gliomas biologically differ from cortical high-grade gliomas (54). Using postmortem tissue and more recently tissue obtained at the time of diagnosis by stereotactic biopsy, gains in platelet-derived growth factor receptor alpha (PDGFRA) and poly (ADP-ribose) polymerase (PARP-1) have been identified as potential therapeutic targets (92; 49; 58). Further work demonstrated focal amplifications of genes within the receptor tyrosine kinase-RAS-phosphoinoside 3-kinase pathway in nearly one half of patients, suggesting that PDGFRA and MET are potential druggable targets (68; 58). This work also suggested that diffuse intrinsic brainstem gliomas had distinct gene expression signatures related to developmental processors different from non-brainstem and low-grade brainstem gliomas.

Subsequently, whole-genome sequencing found that the majority of diffuse intrinsic brainstem gliomas also had mutations of histone H3 (87; 54), suggesting that defects of chromatin architecture is critical in pathogenesis. Diffuse midline glioma, which is defined by loss of H3 K27 trimethylation is due primarily to an H3 K27 M mutation, but may less frequently be due to over expression of EZHIP (13).

Two subgroups of recurrent H3F3A mutations associated with pediatric high-grade gliomas have been identified, confirming disrupted epigenetic regulatory elements. Those with brainstem gliomas predominately have K27 cluster mutations; the same cluster being seen primarily in other midline glioblastoma multiforme tumors of childhood, namely those of the thalamus and spinal cord (82). For this reason, diffuse midline glioma H3 K27M-mutant was incorporated into the 2016 WHO classification schema (52). Both histone H3.3 and H3.1 substitutions are seen (73). The same tumors also have frequent TP53 mutations (82). A recurring activation in the ACVR1 gene, which encodes a type 1 activin receptor serine/threonine kinase has been found on 20% of diffuse brainstem gliomas and may activate the BMP-TGF-beta signaling pathway (83). These latter findings suggest a novel molecular target for diffuse brainstem gliomas. Overall, H.3.3K27M are seen in approximately 80% of diffuse pontine gliomas, with cosegregation of other genes being location dependent, as PDGFRA alterations are most frequent in the pons and FGFR1 in the thalamus. Brainstem gliomas also have CCND2 amplification. H3.3K27M are also associated with ATRX mutations, which are more common in older children (21). H3.1K27M tumors primarily occur more often in the pons, especially in younger patients, than those arising in the H3.3 variant and carry a slightly better prognosis. There is an enrichment of P13K pathway mutations. In H3.3 or H3.1 wild-type cases, there is a great deal of molecular genetic heterogeneity (54). They may be driven by EGFR or PDGFR (54).

Infant brainstem gliomas have the best prognosis, tend to be low-grade, and may harbor NTRK fusions (54). These gene fusions may also involve other tyrosine kinases, including ALK and ROS1 (31; 15). Comparative multidimensional molecular analysis of pediatric diffuse intrinsic pontine gliomas demonstrates relatively distinct subtypes characterized by upregulation of MyC(N-MyC) or Hedgehog (Hh) signaling. Within the Hh subgroup there is upregulation of PTCH (77).

It is unclear whether the histone mutation is a driver of tumor development, and it is likely that other molecular genetic changes are required for tumor development and growth (65; 36). As more tumor tissue has become available, various animal models have been developed; however, most have not been as informative as initially hoped. New technologies, such as in utero electroporation have shown promise in unlocking the significance of the recognized pathohistological heterogeneity of diffuse intrinsic brainstem gliomas and overcoming some of the limitations of previous mouse modeling (67; 80).

Epidemiology

Brainstem gliomas constitute between 10% and 20% of all childhood brain tumors. Brainstem gliomas are the third most common form of posterior fossa tumor of childhood, occurring less frequently than cerebellar astrocytomas and medulloblastomas. Brainstem gliomas can occur at any age during childhood, including the first year of life when they most likely represent congenital tumors. The median age for occurrences is between five and nine years of age. Brainstem gliomas may also occur in adults and by and large tend to be more focal lesions; they are more likely to involve the upper brainstem with contiguous thalamic involvement. There is no clear sex predominance.

Prevention

The risk factors for brainstem gliomas are unknown. There is a relationship of such tumors with neurofibromatosis type 1.

Differential diagnosis

Distinctions between brainstem gliomas of childhood and other posterior fossa tumors are usually not difficult to recognize. The insidious nature of brainstem gliomas and their tendency to sequentially cause progressive cranial neuropathies separates them from the other posterior fossa tumors. Cerebellar astrocytomas are much more likely to present with lateralized cerebellar ataxia and symptoms of increased intracranial pressure early in the course of illness. Similarly, medulloblastomas usually cause midline cerebellar dysfunction associated with headaches and vomiting. Ependymomas may mimic brainstem gliomas but tend to cause less cranial nerve dysfunction and are more likely to present with symptoms and signs of increased intracranial pressure. Rarely, other types of central nervous system malignancies will arise within the brainstem. Both primitive neuroectodermal tumors and medulloepitheliomas may arise within the brainstem and cause cranial neuropathies, ataxia, and long tract signs. In these cases, clinical distinction is difficult. In general, primitive neuroectodermal tumors arise in slightly younger patients and tend to be more localized (91).

Other abnormalities that may mimic brainstem gliomas include brainstem encephalitis, intrabrainstem abscesses, and vascular malformations. Vascular malformations and abscesses tend to present more acutely and cause focal neurologic deficits. Brainstem encephalitis, although extremely rare, usually occurs in the setting of fever and more diffuse symptoms of central nervous system infection. Cerebrospinal fluid analysis may help separate brainstem encephalitis from intrinsic brainstem gliomas; however, because the majority of patients with brainstem tumors do not undergo cerebrospinal fluid analysis prior to diagnosis, these data are often missing. Chronic meningitides, including fungal and tuberculomatous infections of the base of the brain, may result in multiple cranial neuropathies, ataxia, and headaches. Neuroradiographic studies should be able to distinguish between these entities and a brainstem tumor.

On computed tomographic analysis, diffuse intrinsic brainstem gliomas cause diffuse enlargement of the brainstem (51). These tumors tend to hypodense or isodense, and approximately one third will show some degree of contrast enhancement. Various contrast enhancement patterns have been seen, including relatively localized ring enhancement. There has been some suggestion that the tumors that demonstrate ring enhancement may have a more favorable prognosis, but this has not been confirmed in larger series. The upper extent, lower extent, and exophytic nature of brainstem tumors are less than optimally imaged on computed tomography. For these reasons, MRI has supplanted CT as the procedure of choice for the diagnosis and evaluation of patients with brainstem gliomas.

On MR, diffuse intrinsic brainstem gliomas are usually hypointense, with respect to surrounding brain on short TR/TE images and are hyperintense to surrounding brain on long TR/TE images (06). Relatively few tumors are limited to the primary anatomical site, with the majority of tumors involving the pons and at least one other brainstem site. Tumors of the pons are often exophytic, growing ventrally into the prepontine cistern and extending around the basilar artery. Longitudinal infiltration into the midbrain or medulla and axial extension into the middle cerebellar peduncle or cerebellum are common. As was observed in CT studies, enhancement of brainstem gliomas on MR is variable. Once again, various patterns of enhancement have been noted and have not been related to different prognoses.

Tumors of the cervicomedullary junction often have the same neuroimaging characteristics as pilocytic astrocytomas of the cerebellum. On CT, the tumor is usually hypointense and has the same neuroimaging characteristics as diffuse intrinsic brainstem tumors have, without enhancement, on MR. Cervicomedullary tumors may have associated cystic components, and often enhance on both CT and MR.

Tectal masses tend to be localized within the tectal region, with occasional involvement of contiguous mesencephalic regions. These lesions present with associated hydrocephalus. Tectal lesions may contrast enhance on either MR or CT.

Focal pontine lesions are limited to a portion of the pons. They may be cystic, and the solid portions enhance readily on either MR or CT (18).

Diagnostic workup

The diagnostic evaluation of a patient with a presumed brainstem glioma usually begins and ends with MR. This usually displays the extent of the lesion and shows if the tumor is exophytic; MR also shows if there is associated cystic or necrotic change within the tumor. Given the infrequent tendency of primary intrinsic brainstem gliomas to disseminate through the leptomeninges at the time of diagnosis, myelography, and cerebrospinal fluid cytology is rarely indicated. When there is associated hydrocephalus, cerebrospinal fluid cytological examination from the lumbar spaces is contraindicated. Some have recommended MRI of the entire brain and spine, with and without enhancement, for patients with newly diagnosed brainstem tumors. Although the information obtained from these studies can sometimes be helpful, the infrequency of positive results makes it difficult to recommend routine employment of such staging studies.

Neuroimaging findings such as MR spectroscopy using multiparametric imaging (86) and FDG positron emission tomography (93) may be helpful in confirming the presence of a brainstem glioma and have been found to be associated with outcome.

A major unsettled issue in the diagnostic evaluation of patients with brainstem gliomas is whether surgical intervention, except for diversion of cerebrospinal fluid in patients with hydrocephalus, is indicated. Studies have emphasized the specificity of MRI in the diagnosis of brainstem gliomas (02). However, given the grim prognosis, studies are evaluating the safety of stereotactic biopsy and the yield of useful biological information gained from such procedures (76).

Although in adult series other pathologies have been found on biopsy, finding of anything other than a brainstem glioma on biopsy in children with neuroradiographically diagnosed brainstem tumors is infrequent. Given the poor prognosis of patients with brainstem gliomas and the lack of biological information on which to base new therapeutic approaches, there continues to be interest in biopsying diffuse intrinsic tumors to obtain tissue for biological study. However, to date, study of such tissue has not been shown to change management, and, until it does, the rationale for biopsy remains highly problematic. Some prospective studies are using biopsy tissue to guide adjuvant therapy, and this increases the rationale for biopsy, as it may have direct benefit for patients (32; 60).

Some subsets of patients, especially those with cervicomedullary tumors, may benefit from the resection of brainstem glial tumors. However, for patients with diffuse (pontine) lesions, such resections have not been shown to be of therapeutic benefit (02). Biopsies of brainstem tumors have been performed by multiple groups (16; 23). Even after stereotactic procedures, a significant percentage of patients will have increased neurologic dysfunction. The tissue removed at the time of diagnostic biopsy is usually small and may not be representative of the tumor as a whole. Studies performed in the mid-1970s and early 1980s suggested that up to one third of tissue obtained was nondiagnostic. With wider experience with stereotactic means to biopsy the brainstem, surgical morbidity seems somewhat less, and the tissue obtained is more useful for histopathological and biological analysis (76). Despite this, it is unclear how the information obtained at the time of biopsy alters therapy. If the tissue is found to show areas of anaplasia, prognosis is poor. However, in contradistinction, a low-grade biopsy has not been shown to correlate with improved outcome in diffuse intrinsic lesions, either because of sampling error or the tendency for such lesions to dedifferentiate over time. If more effective means become available to treat brainstem gliomas, or if biological study can be shown to more reliably separate patients into definite risk groups, then biopsy of brainstem gliomas may take on more importance.

Management

The most appropriate management for patients with brainstem gliomas remains unsettled. After radiographic diagnosis, most patients are treated without pathologic confirmation (22). However, biopsies at the time of diagnosis are indicated in lesions without classical imaging finding, ruling out brainstem primitive neuroectodermal tumors and other pathology. In addition, stereotactic biopsies may be indicated as part of prospective trials, incorporating molecular-targeted therapies dependent on biopsy results (88; 32; 60). Relatively large, prospective multicentered series have demonstrated that stereotactic biopsies by “experienced” neurosurgeons are relatively safe, as transient neurologic worsening occurs in approximately 10% of those biopsied and permanent neurologic deficits in approximately 5% (32; 60). It is possible that this incidence of surgically related sequalae may increase as the approach becomes more widespread. Reassuring, in the majority of cases, the tissue obtained at the time of biopsy is adequate for extensive molecular analysis (32; 60). Radiotherapy is of at least transient benefit for the majority of patients with diffuse intrinsic tumors. However, conventional doses of radiotherapy, in the range of 5000 to 5500 cGy, cure few, if any, children with diffuse brainstem tumors. There has been interest in increasing the dose of radiotherapy by using hyperfractionated radiation therapy techniques (19; 63; 79; 26). Hyperfractionated radiation therapy is potentially better tolerated by the brain and allows more radiotherapy to be delivered without causing neurologic damage.

Based on these assumptions, different trials have been undertaken over the past 10 years utilizing doses of radiotherapy ranging between 6400 and 7800 cGy in dose fractions ranging between 100 to 126 cGy of radiation delivered twice daily. Despite initial encouraging results in these studies, other multi-institution cooperative trials have disclosed that the majority of children with brainstem gliomas treated at these high doses of radiotherapy will succumb to disease within 18 months of diagnosis (63; 61; 26). Compared to more conventional (once daily) radiotherapy approaches, in trials hyperfractionated radiotherapy has possibly resulted in a higher rate of objective tumor initial shrinkage but has not improved the duration of survival. In one randomized trial, hyperfractionated radiotherapy did not improve event-free or overall survival (55). In addition, nearly one third of patients treated on these higher-dose radiation therapy studies have been steroid-dependent, and intralesional tumor necrosis with associated increased neurologic dysfunction has been noted in nearly 10% of treated patients.

Other radiotherapy techniques have been suggested for children with brainstem gliomas. Hypofractionated radiotherapy trials have shown similar survival rates and toxicity profiles (39; 89; 90; 35).

The published experience with chemotherapy for children with brainstem gliomas is surprisingly scant (63; 27). A variety of drugs, including cisplatinum, carboplatinum, various nitrosoureas, cyclophosphamide, and ifosfamide have been used both singularly and in combination for children with recurrent brainstem gliomas. The results are disappointing, with most series reporting objective response rates of less than 20%, independent of the type of agent or agents used. In one report, a form of interferon, beta interferon, was utilized in nine children with recurrent brainstem gliomas (03). Two patients had a greater than 50% reduction in tumor bulk after treatment, and four others had significant periods of stabilization. However, when beta-interferon was added during and after radiotherapy, no improvement in outcome was found (64). Temozolomide plus 0-6-benzylguanine was not an effective approach (86). Reirradiation was used in the past and is still being used for children with recurrent disease following radiation at diagnosis; some patients have had at least transient benefit (28).

Trials utilizing adjuvant chemotherapy for children with brainstem gliomas have also been disappointing (40). To date, there has been no benefit seen for the use of any type of adjuvant chemotherapy including temozolomide or medtronic regimens, either given prior to or after radiotherapy (25; 41; 10; 48; 78; 17; 14). Despite this fact, given the poor results of available means of treatment, studies have been completed evaluating chemotherapy prior to the initiation of radiation therapy, as an attempt to both identify active agents for patients with brainstem gliomas and improve survival (50; 41). In a COG study of 32 patients with newly diagnosed brainstem gliomas, only 10 + 5% of patients responded to chemotherapy with drug regimens of carboplatin, etoposide, and vincristine, or cisplatin, etoposide, cyclophosphamide, and vincristine. Another study utilizing more intensive pre-irradiation chemotherapy that included intravenous high-dose methotrexate suggested a somewhat higher median survival time but increased toxicity, requiring frequent hospitalizations (24; 30). Another multiagent regimen concurrent and following radiation therapy using thiotepa and antiangiogenic therapy has demonstrated good tolerability and a possible increase in long-term survivors (72).

Similarly, studies are presently underway coupling chemotherapy with radiotherapy in an attempt to increase tumor control, partially by increasing the effectiveness of radiotherapy (42). One study utilizing topotecan during radiotherapy did not demonstrate an improvement in survival (08). Motexafin-gadolinium, a potential radio sensitizer, also did not improve outcome when coupled with radiotherapy (09). Another approach under investigation is utilizing molecular-targeted agents, with radiosensitization characteristics, during radiotherapy, without clear-cut benefit to date (34; 71). Similarly, the addition of temozolomide during and after radiotherapy did not improve survival (05).

Because of the abnormalities in epigenetic regulation at or near critical regulatory histone residues, there is also significant interest in the utilization of drugs affecting chromatin remolding (82; 88; 58).

Other molecular targets are being explored. Nimotuzumab, an anti-EGFR monoclonal antibody, has shown possible benefit in a subset of patients (07; 57). Multiagent molecular-targeted approaches following radiation based on extensive targeted sequencing are actively being employed; such approaches are feasible, but efficacy and safety are still to be shown (47; 60; 58). Immunotherapy approaches, using check-point inhibitors, T-Cell, Car-t-cell, and vaccine studies are underway (58; 56).

Younger children, especially those less than three years of age, may be more responsive to chemotherapy, and, in some, a chemotherapy-alone approach has been successful (11).

In patients with recurrent diffuse intrinsic tumors, reirradiation may be of some benefit (38).

A major limitation of the efficacy of chemotherapy, molecular-targeted therapy, or, likely, immunotherapy is lack of agent delivery to the tumor due to an intact blood-brain barrier in the often nonenhancing, diffuse lesions. To date, means to safely disrupt the barrier have failed. A new approach is coupling microbubble infusion with low-intensity focused ultrasound (LIFU) to transiently and selectively open the barrier in the brainstem, thus, allowing for better agent delivery.

The management for patients with exophytic brainstem gliomas, especially those arising at the cervicomedullary junction, remains somewhat controversial. Some have suggested observation alone, after confirmation of histology by biopsy (29). There have been reports to suggest that such tumors can be extensively resected and are usually pilocytic astrocytomas (20; 81; 45; 46). Short-term follow-up studies suggest that patients, after extensive resection, can be carefully followed without any other forms of specific intervention. However, such resections will result in increased neurologic morbidity in a sizable majority of patients, with some patients becoming respirator-dependent. It is unclear whether the best management of those patients with exophytic cervicomedullary tumors is complete resection or partial resection followed by either radiotherapy or chemotherapy. Because the majority of patients with such lesions harbor low-grade tumors, there is hesitancy to utilize radiotherapy—especially in young patients. There is some information that the use of various drugs, such as the combination of carboplatinum and vincristine, may be effective in patients with recurrent or progressive low-grade brainstem tumors (62).

The management of patients with tectal lesions also is unsettled (85). It is unclear whether such patients require surgical conformation of the pathology of the tumor or can just be followed neuroradiographically. Most observers now suggest that patients with isolated tectal lesions presenting with hydrocephalus can be watched carefully without any specific intervention after cerebrospinal fluid diversion, provided there is no evidence of tumor progression (69; 75). Larger lesions, especially those having a tumor volume of greater than 10 cm3 at presentation, are at higher risk of progressing (84). At the time of tumor progression, the majority of patients have been given local radiotherapy. Long-term disease control for patients with localized midbrain tumors after irradiation is poorly documented; however, in older CT-based studies, patients with such lesions had a relatively favorable prognosis after radiotherapy, with greater than 50% alive and free of progressive disease five years after treatment.

The management of localized pontine lesions is unclear but includes surgical resection alone or biopsy followed by local radiotherapy (18). Juvenile pilocytic astrocytomas may also occur in other regions of the brainstem with similar variable results (44; 46).

At time of recurrence, reirradiation has been used and may result in short-term clinical benefit and some prolongation of survival (38; 04; 12).

Special considerations

Anesthesia

Approximately one third of patients with brainstem gliomas have increased intracranial pressure at the time of diagnosis, and anesthesia techniques for patients with increased intracranial pressure must be utilized. In addition, some patients with brainstem gliomas will have lower cranial nerve dysfunction at the time of diagnosis, and there may be difficulty in weaning patients off artificial ventilation.