Updates to contemporary care of glioblastoma

Abstract

Glioblastoma is the most common and aggressive primary malignant brain tumour in adults. Over the years, the diagnostic criteria have been refined, incorporating molecular markers with histology to reflect the underlying tumour biology. Advances in genomics and molecular techniques are currently focused on trying to subcategorise the tumours for potential prognostic and clinical purposes. Treatment consists of multiple modalities including surgery, radiation, and systemic treatments. While targeted therapy traditionally has not been effective due to the heterogeneity of the tumour, fusions and certain mutations may benefit from specific targeted therapies. Tumour treating fields is a novel treatment that has shown survival benefit in newly diagnosed glioblastoma, and ongoing trials are underway to capitalise on its mechanism. Finally, special consideration should be given to palliative care in light of the high morbidity and mortality associated with glioblastoma. In particular, the balance of aggressive care and quality of life must be weighed in the geriatric population though contemporary studies have shown good tolerance for a hypofractionated course of chemoradiation.

Graphical abstract

Introduction

Glioblastoma is the most common and aggressive primary malignant brain tumour in adults, accounting for nearly half of all malignant brain tumours with an annual incidence rate of 13,000 cases in the United States and an average of 2,530 new cases per year in England.1, 2, 3 Median overall survival (OS) in patients well enough to have aggressive, multi-modality treatment is 15 months, although whole population median survival is seven months due to inability to tolerate aggressive treatment.1, 4, 5, 6 Despite ongoing efforts, this number has not changed significantly over the last few decades.

Glioblastoma incidence increases sharply with age, peaking in adults aged 75–84 years.⁷ Males are nearly 50% more likely than females to be affected with incidence rates of 5.6 per 100,000 in males versus 3.5 per 100,000 in females.7, 8 There is a substantial economic burden associated with glioblastoma. The mean total healthcare costs per patient range from $117,000 to $184,000 for just the first year of treatment.9, 10 With approximately 9,000–9,200 annual glioblastoma deaths in the United States between 2018 and 2022, considerable progress is needed to refine treatment and improve clinical outcomes for this devastating disease.1, 11

Surgical care

Maximal safe resection remains the cornerstone of modern surgical care for glioblastoma, with particular emphasis on preservation of neurological function. A meta-analysis of 37 studies showed extent of resection, namely gross total resection, correlated with improved survival.¹² Advanced preoperative MRI with tractography and stereotactic image guided neuronavigation, including intraoperative MRI, is frequently utilised to safely maximise extent of resection. Achievement of gross total or supra-marginal resection can also be enhanced by use of intraoperative aminolevulinic acid hydrochloride (5-ALA), a substance that causes glioma cells to fluoresce under blue light, improving surgical delineation between tumour and white matter. Though there is little consensus of the use of 5-ALA on overall or progression-free survival, its use has been associated with an up to 26% higher rate of gross total resection over conventional surgery.13, 14, 15 A systemic review and meta-analysis comparing effectiveness of intraoperative MRI and 5-ALA for achieving gross total resection showed that while techniques were superior to conventional neuronavigation, neither one modality was superior compared to the other.¹⁶ Stereotactic biopsy is typically reserved for multifocal disease, tumours involving eloquent areas like the brainstem or thalamus, or patients with poor functional status.¹⁷

Laser Interstitial Thermal Therapy (LITT) has emerged as a minimally invasive surgical option for patients who are not candidates for craniotomy. The technique involves stereotactic placement of an MRI-compatible laser probe into the tumour core, which then causes coagulative necrosis of the targeted tissue.18, 19 LITT achieves high ablation rates with a median of 94.8% in IDH-wildtype GBM and can effectively reduce tumour volume.20, 21 Although clinical outcomes of LITT have been promising in select patient populations, it is not without risks, including cerebral oedema, haemorrhage, and permanent neurological deficits.22, 23 Gliadel wafers (Arbor Pharmaceuticals, Atlanta, GA) are bio-degradable polymers placed in the resection bed at the time of surgery to provide controlled release of carmustine, an alkylating nitrosurea, though it’s use and benefit remain debated due to minimal randomised clinical evidence and statistically significant but clinically modest overall survival benefit in multiple cohort studies (with no consistent impact on progression-free survival).24, 25 Focused Ultrasound (FUS) represents a rapidly evolving field that includes thermal ablation and non-thermal blood-brain barrier (BBB) modulation to improve penetration of systemic chemotherapeutic agents.²⁶

Pathology

Historically, histopathology formed the basis for glioma diagnosis and grading. Traditional criteria include anaplasia, hypercellularity, nuclear pleomorphism, pseudopallisading necrosis, and microvascular proliferation. In 2016, the World Health Organization (WHO) updated its classification for central nervous system (CNS) tumours (CNS4) by combining histological and molecular information into an integrated diagnosis to improve diagnostic accuracy and prognosis and the fifth edition continued with changes that further advanced the role of molecular diagnostics in tumour classification.27, 28 CNS5 simplified adult gliomas from 15 entities to three types: Astrocytoma, isocitrate dehydrogenase (IDH)-mutant; Oligodendroglioma, IDH-mutant and 1p/19-codeleted; and Glioblastoma, IDH-wildtype, with only IDH wild-type being defined as glioblastoma based on molecular data, regardless of histological findings.²⁸ The previously designated ʻglioblastoma, IDH mutant,’ is now reclassified as grade 4 astrocytoma, IDH mutant. While the IDH1 R132H mutation is the most common mutation in IDH mutant gliomas, there are other mutations that can occur in the IDH1 and, less commonly, the IDH2 gene. Thus, the authors recommend verifying IDH wildtype status with next generation sequencing in patients 40 and younger to rule out noncanonical IDH mutations. The definition of glioblastoma was also further expanded from beyond the histological parameters to include the presence of one or more of three genetic parameters – telomerase reverse transcriptase (TERT) promoter mutation, epidermal growth factor (EGFR) amplification, combined gain of chromosome 7 and loss of chromosome 10 – as sufficient to render a diagnosis of glioblastoma, IDH wild-type, in the absence of the traditional histological findings.28, 29, 30 Finally, CNS 5 distinguishes between adult and paediatric high grade gliomas based on distinct molecular profiles.²⁸

There has been considerable translational work to improve characterisation and elucidate the heterogeneity implicit to glioblastoma. The Cancer Genome Atlas Research Network (TCGA) catalogued genomic abnormalities to describe the mutational spectrum of glioblastoma.³¹ Single cell RNA sequencing has highlighted that glioblastoma displays significant intratumor heterogeneity with different subtypes within the same tumour, which may underlie treatment failure and disease recurrence.32, 33

More recently, global DNA methylation profiling, which looks at epigenetic changes, has been used to stratify glioblastoma, with potential clinical and therapeutic implications.34, 35, 36, 37 Methylation array may be an important tool to highlight changes associated with long term survival, implicating pathways involved in long term survival.³⁸

Liquid biopsy

Liquid biopsy is an emerging tool for the noninvasive molecular diagnosis and longitudinal monitoring of GBM, helping to address the known limitations of tissue sampling and conventional imaging to reliably distinguish progression from treatment effect. A RANO taskforce reviewed the principal analytes (circulating tumour DNA, circulating tumour cells, extracellular vesicles) in blood and cerebrospinal fluid and have proposed clinical applications for diagnosis when tissue is unavailable, detection of residual disease, diagnosis of pseudoprogression, and prognostication.³⁹ Because the blood-brain barrier can limit the amount of tumour-derived cell-free DNA in plasma, CSF is currently the most informative biofluid: pooled data show ctCNA detection in about 82% of CSF samples versus about 16% of plasma, with around 90% molecular concordance with tumour tissue.⁴⁰ Advances in ultrasensitive sequencing are starting to enable new detection of GBM-associated ctDNA in blood in the absence of prior tissue diagnosis.⁴¹ Standardisation of biofluid collection, analyte selection, and detection methodology, along with prospective multicenter validation, remains necessary prior to incorporating liquid biopsy into routine clinical practice.

Imaging

The assessment of therapeutic response in glioblastoma has historically been one of the most contentious and technically demanding tasks in neuro-oncology. The Response Assessment in Neuro-Oncology (RANO) criteria are imaging guidelines designed to standardise the process of assessing treatment response in clinical trials for gliomas. Unlike preceding criteria, it addressed the concept of pseudoprogression, the false assumption of tumour progression in setting of treatment associated inflammatory changes.⁴² RANO criteria stipulated that disease progression could only be identified within 12 weeks of completion of radiotherapy if the new enhancement was outside the radiation field or if there was histopathological confirmation of viable tumour.43, 44 RANO criteria also integrated T2-weight and fluid-attenuated inversion recovery (FLAIR) imaging into the assessment process. This helped address the issue of false treatment response, or pseudoresponse, which was becoming increasingly prevalent with the emergence of antiangiogenic agents like bevacizumab.43, 45, 46

The 2023 publication of the RANO 2.0 criteria marks the most significant refinement in neuro-oncology imaging assessment to date.⁴⁴ It provides a single unified framework for all adult gliomas and includes their integrated molecular diagnosis. One of the most important change was the mandate of the first post-radiotherapy MRI (typically obtained three to five weeks after completion of radiotherapy) as the primary baseline for response assessment in patients receiving radiation.42, 47 This change significantly improved the quality control for baseline imaging, making it arguably the most significant of all the updates in RANO 2.0 for establishing a reliable assessment of tumour response. RANO 2.0 criteria also uses the 2D measurements from previous guidelines and combines it with 3D volumetrics, which is particularly advantageous for glioblastoma, which often exhibit irregular, non-spherical growth patterns around resection margins that are poorly captured by two linear measurements.

Treatment

Radiation

Radiation has been a long-standing cornerstone of glioblastoma therapy by offering microscopic and macroscopic local control. Multiple studies have shown survival benefit with the addition of radiation therapy after surgery compared to chemotherapy alone or best supportive care.48, 49, 50, 51

Standard of care for newly diagnosed glioblastoma includes maximal safe surgical resection, followed by concurrent TMZ with partial brain irradiation and adjuvant TMZ.⁵² Multiple studies have shown improved survival using standard fractionation (1.8–2 Gy per day) to 60 Gy.49, 53 Dose-escalation, dose intensification, accelerated fractionation, hypofractionated boost, and stereotactic boost, have not shown any clear benefit in the glioblastoma population.53, 54, 55

Currently, there is no clear survival advantage with proton beam radiation versus conventional photon beam radiotherapy.⁵⁶ Proton beams have a steeper dose gradient and theoretically allow for more focused delivery with substantially reduced dose exposure beyond the target volume.⁵⁷ One may consider proton therapy if vulnerable regions such as the brainstem or optic chiasm are within the margins of the radiation field to minimise collateral damage. While there is no significant difference in time to cognitive failure between proton and photon therapy, there was reduction in fatigue and TMZ-associated myelosuppression for patients.⁵⁶

Given the infiltrative nature of glioblastoma, focused radiation such as stereotactic radiosurgery (SRS) has traditionally been reserved for specific cases of focal recurrences. Interestingly, a recent retrospective study compared a five fraction SRS course to a traditional 30 fraction course and 15 fraction course and found comparable survival outcomes, though there were fewer patients in the SRS arm and there were a higher proportion of gross total resections in the SRS arm.⁵⁸ Brachytherapy is a modality that utilises implantation of radioactive vectors within the tumour cavity at the time of resection. GammaTile is a novel brachytherapy approach that is FDA-cleared for recurrent brain tumours or newly diagnosed malignant brain tumours, including glioblastoma, with multiple trials ongoing.59, 60 Intraoperative radiotherapy using an external radiotherapy source and applicator has been assessed in the phase 3 INTRAGO-2 trial, which appeared to show no benefit from the addition of intraoperative radiation.⁶¹

A recurrent challenge faced by glioblastoma patients who receive radiation is the issue of radiation necrosis and pseudoprogression. Radiation necrosis is a severe and more permanent local tissue reaction to radiotherapy and can look very similar to pseudoprogression, which is an increase in contrast enhancement on MRI mimicking tumour progression but will eventually spontaneously improve.⁴⁴ Unfortunately, distinguishing between tumour progression and pseudoprogression via conventional anatomic MRI is quite problematic with pooled sensitivity and specificity of 68% and 77% respectively.⁶² Advanced imaging including perfusion, diffusion, amino acid imaging may improve our ability to prospectively identify pseudoprogression cases.⁶³ Patients with MGMT methylated tumours are at higher risk of pseudoprogression.⁶⁴ Symptomatic patients often require steroids. Bevacizumab, an antiangiogenic agent, may be employed if prolonged steroids will be required.⁶⁵ Unfortunately, early use of bevacizumab to control pseudoprogression may preclude future trial enrolment as bevacizumab is often an exclusionary criteria for recurrent glioblastoma trials.

Systemic treatment

Unfortunately, there are limited effective systemic treatments for glioblastoma due to the inherent heterogeneity of the tumour and limited penetration of drugs across the blood brain barrier. Concerted efforts have been focused on improving efficacy of systemic treatments (Table 1, Table 2). While the ultimate hope is for curative treatment, current focus is on delaying progression and improving survival.

Table 1.

Conventional chemotherapy for glioblastoma.

Drug name	Biologic action	Setting	Key notes
Temozolomide (TMZ)	Induces double-stranded DNA breaks via DNA mismatch repair and base excision repair; also functions as a radiosensitizer	Newly diagnosed; some recurrent use	Addition to radiotherapy is considered standard of care.52 Silencing of MGMT via methylation of the promoter is associated with improved response.
Lomustine (CCNU)	Alkylates both DNA and RNA and can cross link DNA	Recurrent disease	Intensification of alkylating therapy has been explored by combining nitrosurea compounds, such as lomustine, with TMZ66, 67, 68
Bevacizumab	Humanised monoclonal antibody against vascular endothelial growth factor (VEGF)	Recurrent disease	Initial phase 2 studies showed significant radiographic response and improved progression free survival.69, 70, 71 Failed to demonstrate improvement in overall survival.72, 73

These are commonly used first and second line systemic therapies for glioblastoma regardless of molecular profiling.

Table 2.

New targeted therapies for glioblastoma.

Drug name	Biologic action	Setting	Key notes
Erdafitinib	Fibroblast growth factor receptor (FGFR) inhibitor	High and low grade gliomas with FGFR fusion	The RAGNAR study, which included heavily pretreated glioblastoma patients, showed poor response rates.74
Larotrectinib	Tropomyosin receptor kinase (TRK) inhibitor	CNS tumours with neurotrophic tyrosine kinase (NTRK) fusions	Case report and small studies of responses to TRK inhibition.75, 76, 77, 78
Entrectinib	Tropomyosin receptor kinase (TRK) A/B/C, ROS proto-oncogene 1 (ROS1), anaplastic lymphoma kinase (ALK) tyrosine kinase inhibitor	CNS tumours with NTRK aberrations	The STARTRK-NG trial demonstrated response.79
Dabrafenib	V-raf murine sarcoma viral homologue B1 (BRAF) inhibitor	High and low grade gliomas with BRAFv600E	Several case reports of response to dabrafenib. The ROAR basket trial showed clinically meaningful activity with dabrafenib and trametinib.80, 81, 82, 83, 84
Trametinib	Mitogen-activated extracellular signal-regulated kinase (MEK) inhibitor

While targeted therapies traditionally have been ineffective glioblastoma due to their inherent heterogeneity, there are specific cases in which they may benefit the patient. Typically, fusions are more attractive targets as they often function as a key driver for tumour growth. BRAF mutations have also been show to play a key role in initial tumorigenesis and may benefit from targeted agents.

Temozolomide (TMZ)

TMZ is an oral cytotoxic alkylating agent that induces double-stranded DNA breaks via DNA mismatch repair and base excision repair.85, 86 One of the key sites of action is the O6 site of guanine creating O6-methylguanine. While it constitutes only a minority of adducts formed by TMZ, it plays an important role in the cytotoxic action of the drug by utilising the mismatch repair (MMR) mediated futile cycling.⁸⁷ MGMT is the only mechanism to repair these adducts and reduction of levels of the enzyme is critical to the cytotoxic potential of TMZ. Clinically, silencing of MGMT via methylation of the promoter is associated with improved response to TMZ.⁸⁸

The landmark study by Stupp et al., showed that the addition of TMZ to radiotherapy compared to radiation alone improved survival and is now considered standard of care for newly diagnosed glioblastoma patients.⁵² A phase 2 study (GEINO 14–01) compared six cycles of adjuvant TMZ to 12 cycles and found no improvement in overall survival or progression free survival regardless of MGMT status.⁸⁹

Recognition that decreased MGMT activity may increase sensitivity and response to TMZ has led to different strategies to try to capitalise on that knowledge. A phase 2 study of dose dense TMZ (21 days on, seven days off) for newly diagnosed glioblastoma did not show increased efficacy with this alternative regimen.⁹⁰ A phase 3 study of newly diagnosed malignant glioma in the elderly showed that dose dense TMZ (seven days on, seven days off) was noninferior to radiotherapy alone in MGMT methylated patients, providing an alternative treatment to radiotherapy.⁹¹ Caution is warranted with excessive TMZ exposure as there is a risk for TMZ induced hypermutation, which is associated with worse prognosis and shortened survival.⁹²

TMZ is relatively well tolerated. The most common side effects include fatigue, myelosuppression, nausea and vomiting. In addition, prolonged chemotherapy with alkylating agents is associated with risk of myelodysplastic syndromes and secondary leukaemias, although the risk of these in patients with GBM is limited due to their poor survival.⁹³

Lomustine and carmustine

Lomustine (CCNU) is an alkylating agent in the nitrosourea family. It alkylates both DNA and RNA and can cross link DNA.⁹⁴ It also acts on O6-guanine forming O6-chloroethylguanine, which can be reversed by MGMT. Historically, it has been used in recurrent glioblastoma with OS of six to nine months with the majority of responders being MGMT methylated.⁹⁵ It is often used as the control arm of trials with the majority of the trials showing lomustine to be as effective as the experimental treatment.96, 97 Intensification of alkylating therapy has been explored by combining nitrosurea compounds, such as lomustine, with TMZ.66, 67, 68 As both drugs affect different types of DNA damage, there is potential for synergistic or additive effects. CeTeG/NOA 09 looked at the addition of lomustine to TMZ with radiation in MGMT methylated newly diagnosed glioblastoma patients.⁶⁶ While the study was small (n = 129), there was significantly improved survival in the combination arm. This is being explored in a larger population in the NRG BN011 trial.

Side effect profile with lomustine is similar to TMZ. Fatigue, myelosuppression, nausea and vomiting are common. Unlike TMZ, lung toxicity including pulmonary fibrosis is a rare but severe side effect of lomustine, thus baseline pulmonary function tests (PFTs) may be requested are obtained prior to initiation of the treatment.⁹⁸

Bevacizumab

Because GBM is a highly vascular tumour with high expression of vascular endothelial growth factor (VEGF), antiangiogenic drugs seem to be a promising therapeutic strategy.⁹⁹ Initial phase 2 studies showed significant radiographic response and improved progression free survival resulting in accelerated approval for bevacizumab, a humanised monoclonal antibody against VEGF.69, 70, 71 Two subsequent phase 3 trials of bevacizumab combined with the Stupp protocol failed to demonstrate improvement in overall survival.72, 73 Other phase 3 studies with antiangiogenic agents for recurrent glioblastoma also failed to show survival benefit.96, 100

While bevacizumab and many other antiangiogenics do not improve survival, they do improve progression free survival.72, 73, 101, 102 Specifically, bevacizumab is effective in controlling oedema associated with treatment including radiation.¹⁰³

Targeted therapies

There are frequently recurring mutations seen in many glioblastoma specimens. This includes epidermal growth factor (EGFR) mutations, including EGFRvIII, and amplifications, cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) homozygous loss, neurofibromin 1 (NF1) mutation, and platelet-derived growth factor receptor alpha (PDGFRa) mutation.104, 105 Less common mutations include v-raf murine sarcoma viral homologue B1 (BRAF) mutations (including BRAFv600E), fibroblast growth factor receptor (FGFR) mutations, overexpression, or fusions, and neurotrophic tyrosine receptor kinases (NTRK) fusions. All of these create potential attractive targets for therapies. Unfortunately, the heterogeneity inherent to the tumour often renders this approach ineffective.

Multiple studies have tried to target EGFR with very little success.106, 107, 108 Part of the challenge involves poor blood brain barrier penetration rendering anti-EGFR monoclonal antibodies or antibody-drug conjugates ineffective.¹⁰⁹ In addition, selective pressures within the tumour microenvironment may influence occurrence or disappearance of EGFR mutations during tumour development.110, 111 Similarly, targeting somatic NF1 mutations or CDKN2A/B deletions, both of which affect the Ras/rapidly accelerated fibrosarcoma (RAF)/MEK pathway, have not been successful likely related to the heterogeneous landscape in the tumour.¹¹² Regorafenib, a multikinase inhibitor, initially showed survival benefit in recurrent glioblastoma in the phase 2 REGOMA study but follow up phase 2/3 studies of its use in newly diagnosed and recurrent glioblastoma showed no superiority of the drug over either lomustine or temozolomide.97, 113 There may be a few circumstances in which targeted drugs may be effective (Table 2). For example, gene fusions are genetic alterations that derive from chromosomal rearrangements producing fusion proteins that drive tumour growth and development.¹¹⁴ The RAGNAR study, which included heavily pretreated glioblastoma patients, showed response rates in high and low grade gliomas.⁷⁴ Neurotrophic tyrosine receptor kinase (NTRK) fusions are another potential therapeutic target. There have been case reports and small studies of responses to TRK inhibition.75, 76, 77, 78 The STARTRK-NG trial included CNS tumours with NTRK aberrations and demonstrated response to entrectinib.⁷⁹ Finally, while BRAFv600E mutations are rare in glioblastoma, there have been several case reports of response to vemurafenib and dabrafenib.80, 81, 82, 115 The ROAR basket trial included high and low grade gliomas and showed clinically meaningful activity with dabrafenib and trametinib in tumours with BRAFv600E mutation.⁸³ Ongoing investigation is still underway to define the select tumours that benefit from targeted drugs.

Given the inherently poor prognosis associated with glioblastoma, there is recognition that clinical trials play a significant role with National Comprehensive Cancer Network (NCCN) guidelines recommending consideration of clinical trials for recurrent glioblastoma patients.¹¹⁶ Unfortunately, enrolment of patients in primary brain tumour trials are low with only 8–11% of newly diagnosed glioblastoma patients enroling in clinical trials.¹¹⁷ There is increasing awareness of flaws in historic clinical trial designs with efforts to modernise the approach with inclusion of adaptive design to enrich patients with specific biomarkers to appropriate treatment arms.¹¹⁸ However, updating clinical trial design is only part of the answer. Recognition of enrolment barriers is also critical for successful clinical trial recruitment; for instance, travel more than 1 h from the home significantly reduces the ability of willingness to enrol in a trial.¹¹⁹ Increased partnership with community neuro-oncology providers and larger academic centres may help with decentralisation of trials to improve and equalise the opportunity for trial participation.

Tumour treating fields (TTF)

Tumour-treating fields (TTF) is a relatively novel anti-cancer therapy using a noninvasive device to pass an alternating medium frequency (150–200 Hz), low voltage current through the tumour via arrays placed on the skin. This interferes with multiple biological processes including mitosis, autophagy, DNA damage response, cell motility, blood brain barrier permeability, and anti-tumour immunity.120, 121, 122

In a phase 3 trial comparing TTF with the Stupp protocol to Stupp protocol alone, both progression free and OS were improved in the experimental arm, with an improvement in median survival from 16 to 20.9 months. Post hoc analysis of elderly patients also showed improved survival.¹²³ This study established a threshold of an average of at least 18 h/ day (75%) to receive the survival benefit from TTF. Because main toxicity is skin irritation, dermatologic care is essential to maximise usage by maintaining skin health, and multiple recommendations have evolved to help minimise dermatologic adverse events.124, 125, 126 Recent data from the phase 3 TRIDENT study that looked at use of TTF with chemoradiation and adjuvant temozolomide to chemoradiation and adjuvant temozolomide with TTF reported no increase in overall survival for patients who initiated TTF at the start of chemoradiation.¹²⁷ A study of TTF with checkpoint inhibition (EF-41/KEYNOTE D58) in newly diagnosed glioblastoma just finished enrolment with results pending.

Immunotherapy

Despite the transformative impact of immune checkpoint blockade in other solid tumours, immunotherapy has not yet yielded meaningful survival gains in GBM, although substantial work continues in this space. Two large randomised phase III trials of anti–PD-1 therapy in newly diagnosed tumours were negative: CheckMate 548 added nivolumab to radiotherapy and temozolomide in MGMT-methylated tumours and showed no improvement in progression-free or overall survival, while CheckMate 498 compared nivolumab plus radiotherapy against temozolomide plus radiotherapy in MGMT-unmethylated tumours and similarly found no improvement in OS.128, 129 Against this backdrop, the 2-THE-TOP phase II study – combining pembrolizumab with tumour treating fields and temozolomide after chemoradiation –was positive, reporting a median progression-free survival of 12.0 months versus 5.8 months and median overall survival of 24.8 months versus 14.7 months relative to a matched EF-14 historical cohort.130, 131 These encouraging results have motivated an ongoing phase III pembrolizumab study to prospectively confirm the benefit of adding checkpoint blockade to tumour treating fields.

Beyond checkpoint inhibitors, several novel immunotherapeutic platforms are under active investigation. Chimeric antigen receptor (CAR) T-cell therapy, which can be delivered locoregionally and directed against IL-13Rα2, EGFR/EGFRvIII, and more recently B7-H3, has demonstrated feasibility, acceptable safety, and early signals of antitumor activity in multiple phase 1 trials for recurrent glioblastoma, though durable responses remain uncommon.132, 133, 134 Tumour-antigen and personalised neoantigen vaccines likewise generate measurable intratumoral and circulating T-cell responses and have shown promising survival signals in early-phase studies.135, 136, 137

Supportive care

Due to the aggressive nature of the tumour, supportive and palliative care is a key component to contemporary glioblastoma care. Early involvement of palliative care has been shown to improve not only quality of life but also survival in oncology patients.¹³⁸

In neuro-oncology, symptoms arise from the tumour and from treatment. The updated Society for Neuro-oncology and European Association of Neuro-oncology guidelines for palliative care in malignant glioma patients reported cognitive changes, seizures, fatigue, headache, and pain as the predominant issues for patients.¹³⁹ Steroids are commonly used for temporary relief with bevacizumab used for those patients intolerant of steroids.

As a progressive brain disease, glioblastoma patients face the unique challenge in that their disease interferes with their ability to make their own decisions or express their wishes. Half of glioma patients report difficulties with treatment decisions early in their disease course, and this number increases towards the end of life.140, 141 Thus, early assessment and discussion of palliative care is imperative for advance care planning. This may be done within the neuro-oncology team, but at times may benefit from formal palliative care referral, depending on the patient, their symptoms and supportive care needs. Key questions include where patients wish to receive end-of-life care, which is highly dependent on personal, cultural and socio-economic factors. While glioblastoma patients generally are more likely to die at home than the general population, there are factors that influence the place of death in malignant glioma patients.¹⁴² Thus, early involvement with palliative care is imperative for advance care planning and to ensure patient wishes are known and upheld throughout the disease.

Treatment in the elderly

Elderly patients (≥65 years) warrant special consideration. While maximal safe resection remains the initial cornerstone of management, the decision to pursue standard chemoradiation, abbreviated chemoradiation, radiotherapy alone, chemotherapy alone is highly individualised based on patient fitness, comorbidities, and tumour biology.

NOA-08 showed that dose dense TMZ was noninferior to standard radiation (60 Gy) in elderly or frail patients with OS of 8.6 months for temozolomide versus 9.6 months for radiotherapy.⁹¹ Subgroup analysis demonstrated that MGMT-methylated tumours experienced significantly improved survival with TMZ compared to unmethylated tumours and MGMT unmethylated tumours derived greater benefit from radiotherapy. These findings, together with favourable toxicity and treatment-burden profiles, support TMZ monotherapy as a reasonable and often preferable option for frail patients with MGMT-methylated tumours.

A hypofractionated course of chemoradiation has emerged as a meaningful alternative treatment for elderly patients. The Nordic study compared standard radiotherapy (60 Gy) to TMZ alone and hypofractionated radiation (34 Gy in 10 fractions).¹⁴³ This study and subsequent updates reinforced that short-course radiotherapy regimens (e.g., 34 Gy in 10 fractions or 40 Gy in 15 fractions) are not only noninferior but also superior in practicality and tolerability compared with standard radiotherapy.¹⁴⁴ Perry et al. compared hypofractionated radiotherapy with chemotherapy to hypofractionated radiotherapy alone and showed addition of temozolomide improved survival, especially in MGMT methylated patients.¹⁴⁵ Currently, a short course of chemoradiation has become standard of care for the elderly with consideration of temozolomide alone for MGMT methylated patients with poor performance status.

Conclusion

Over the last decades, important advances have improved our understanding of the tumour biology and molecular pathogenesis of glioblastoma. Unfortunately, this knowledge has yet to be translated to significantly improved outcomes for the patients. Treatment remains multidisciplinary with novel therapeutic strategies being developed from preclinical work. Given the overall poor prognosis associated with the disease, incorporation of palliative care is recommended early in the disease course. Elderly patients require special consideration in order to balance treatment with quality of life.

CRediT authorship contribution statement

Matt Williams: Writing – review & editing, Writing – original draft, Data curation, Conceptualization. Sara Elsharkawy: Writing – original draft. Christine Lu-Emerson: Writing – review & editing, Writing – original draft, Supervision, Project administration, Data curation, Conceptualization. Bernadeth Lyn C. Piamonte: Writing – review & editing, Writing – original draft, Investigation, Data curation. Wiley Gillam: Writing – original draft, Data curation. Megan E.H. Still: Writing – review & editing, Writing – original draft, Supervision, Data curation. Ashley P. Ghiaseddin: Writing – review & editing, Writing – original draft, Data curation.

Consent for publication

All details in the clinical vignettes are fictional. All identifying information has been removed from the images in Figures 1 and 2 therefore formal consent was not required.

Funding

This research/review did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Ashley Ghiaseddin reports a relationship with Alexion Pharmaceuticals Inc that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Deciphera Pharmaceuticals that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Novocure Inc that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Nuvation Bio, Inc that includes: consulting or advisory. Ashley Ghiassedin reports a relationship with Servier Medical Affairs that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Telix Pharmaceuticals Limited that includes: consulting or advisory. Christine Lu-Emerson reports a relationship with The Sontag Foundation that includes: travel reimbursement. Christine Lu-Emerson reports a relationship with National Institutes of Health National Cancer Institute that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Monteris Medical Corporation that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with Neosoma that includes: consulting or advisory and equity or stocks. Ashley Ghiaseddin reports a relationship with SpringWorks Therapeutics that includes: consulting or advisory. Ashley Ghiaseddin reports a relationship with CRISPR Therapeutics Inc that includes: equity or stocks. Ashley Ghiaseddin reports a relationship with Prime Medicine Inc that includes: equity or stocks. Matt Williams reports a relationship with Cosyne that includes: consulting or advisory. Matt Williams reports a relationship with Toxiproof.AI that includes: consulting or advisory. Matt Williams reports a relationship with Pear Bio that includes: equity or stocks. Matt Williams reports a relationship with Novocure Inc that includes: funding grants and speaking and lecture fees. Matt Williams reports a relationship with Novartis that includes: non-financial support. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.