Understanding Glioblastoma

What makes Glioblastoma particularly deadly?

The combination of the following features can be seen as root cause for the particularly deadly nature of GBM:

• GBMs contain a subpopulation of cells with stem cell-like properties. Such cells are strong in self-renewal, possess an enhanced DNA-damage repair system, and the ability to differentiate. There is ongoing discussion in the literature as to whether these cells are true stem cells or merely possess stem cell-like properties. Regardless, there is compelling evidence that this subpopulation is a strong driver behind the aggressiveness of the tumor, the high frequency of relapses, and resistance to therapies [1,2,3,4,5].
• GBMs are so-called "cold" tumors for the immune system. They show a low to moderate mutational burden and downregulated MHC expression resulting in a low number of neo-antigens/ neo-epitopes presented on their surface which makes them hard to detect and fight for the immune system [6,7,8].
• GBMs create an immunosuppressive micro-environment by e.g. producing inhibiting metabolites such as lactic acid and overexpressing checkpoint ligands [9,10,11].
• Tumor associated M2-type macrophages account for 30-50% of the GBM mass. These macrophages suppress T-cell activity by secreting immunosuppressive cytokines (TGF-β, IL-10). They also secrete EGF and VEGF, stimulating proliferation and angiogenesis. Finally, they help facilitate tumor growth in hypoxic micro-environments [10,12,13].
• GBMs are highly heterogeneous tumors in which different tumor regions can differ in terms of cell type, genetics, epigenetics, transcriptomics, metabolism and tumor-micro-environment. This heterogeneity makes exact characterization very challenging.
• In case of a relapse, local heterogeneity is compounded by a hard to predict transformation along the time axis. The variety of mutations and transformations statistically increases with size and is exacerbated by selective treatments that control or eliminate only certain cancer subgroups, thereby promoting all other untreated cancer subgroups through selection. Even if the tumor has initially been characterized after surgical removal using state-of-the-art analytics, strong mutations with significant changes in metabolism can alter the nature of the tumor within weeks up to an extent that would require a new analytical characterization, followed by an adjustment of the therapy. This, however, is highly unrealistic, as it would require quarterly brain surgery, which would have a major impact on quality of life. Liquid biopsy methods targeting, for example, circulating tumor cells, circulating DNA/RNA or extracellular vesicles are under development and may help to resolve or at least ease this issue in the medium term. However, they are not yet implemented in Standard of Care.
• GBMs can form internal networks that promote tumor progression and resistance. Recent findings on the network structure of glioblastoma have highlighted key mechanisms such as tumor microtubules and neuron-glioma synaptic connections [14, 15, 16].

The combination of properties listed above makes every GBM unique and treatment incredibly challenging [4,10,17,18].

Why is the standard therapy according to treatment guidelines insufficient?

The Standard of Care for first-line treatment of GBM, according to the German Commission for Diagnostics and Treatment Guidelines (Kommission Leitlinien der Deutschen Gesellschaft für Neurologie) [19] is a 20-year-old radiochemotherapy protocol (initial radiation combined and followed by repeated cycles of temozolomide (TMZ)). This protocol neither differentiates between cell types or mutational patterns, nor does it consider differences in tumor metabolism or tumor microenvironment. Its chemotherapeutic agent TMZ is considered ineffective for the large subgroup of patients with an unmethylated MGMT promotor (approximately 50% of all patients) [21]. Conversely, for the unmethylated subgroup, TMZ, due to its toxicity, can worsen the patient’s general condition, negatively impact quality of life, and can trigger hypermutation [18, 22].

Even if the standard therapy initially works, its efficacy dramatically decreases after just a few months for almost all patients- the relapse is the norm. After 7-8 months, roughly 50% of all patients experience a relapse. After 15 months, 90% of all patients are facing a relapse [23, 24, 25, 26, 27]. The causes for the fast reduction in efficacy are manifold and can be attributed to the unique characteristics of the GBM. The main mechanism of action of both radiation and TMZ is to damage the DNA of tumor cells, thereby stopping replication. However, the destructive effect of therapy is countered by the DNA repair capacity of cells. The lower the grade of methylation of the MGMT promotor, the greater the self-repair capability, which explains why TMZ shows no effect in truly unmethylated patients. Other factors driving resistance include the ineffectiveness of radiation in hypoxic tumor regions and the ability of the tumor to hide from and suppress the immune system [5, 22, 23, 24, 28, 29].

If initially successful, TMZ cycles can be complemented by the application of Tumor Treating Fields (TTFields) which can extend Progression-Free Survival (PFS) by 3 months and Overall Survival (OS) by 5 months [20]. For younger patients with a methylated MGMT-promotor, lomustine is recommended in addition to radiochemotherapy. In case of a relapse, the guideline does not provide specific recommendations but generally lists options such as re-surgery, re-radiochemotherapy, targeted therapies, bevacizumab, and immunotherapies, all preferably within the context of clinical trials.

Radiochemotherapy and TTFields are the only two therapies that have reached their endpoints in phase III clinical trials and thus received EMA approval for GBM. Although there is a plethora of experimental therapies, ranging from chemotherapies and oncolytic viruses to immune therapies such as personalized vaccines or CAR-T, none of these therapies have been able to successfully pass phase III clinical trials. Over the past 25 years, around 350 entities or therapy concepts have been tested globally in around 1,000 clinical trials. Less than 1% were approved [30, 31, 32, 33, 34, 35].

Why is the number of newly approved therapies so low?

From our perspective, the disappointing track record of clinical trials over the past 25 years may, amongst others, also be attributed to the following three factors:

• Firstly, patients are admitted to studies and stratified within studies based on a set of criteria which may be incomprehensive. Considering criteria such as disease stage, MGMT-promoter methylation, Karnofsky Performance Index or a few DNA mutations, the design of today's clinical trials may not reflect the multi-dimensional complexity of a GBM tumor. This may result in an insufficient match between the tested therapy and the true nature of the tumors being treated in the trial group.
• Secondly, data used for patient stratification almost always stems from the time of the most recent surgery, when a tumor sample was obtained and analyzed. The time lag between obtaining the sample and the start of the trial can be significantly longer than the time it takes for the tumor to undergo major transformation. If a therapy has been applied between tumor characterization and the start of the trial, the value of the initial data set is even further reduced. Data older than 3 months and data obtained before a prior therapy may have limited predictive value.
• Lastly, clinical trials typically test mono-therapies, whereby single drugs or medical devices are compared to the Standard of Care. While this approach allows for a scientifically correct evaluation of the therapy candidate, the chances of curing the disease or reaching the study endpoint are low as the approach probably does not address the complex and multi-faceted nature of GBM. Even if the test candidate may be successful in fighting one type of cell or inhibiting one type of metabolism as intended, the tumor may continue its growth as other cells or other metabolic pathways could remain untouched. At the contrary - under these conditions, tumor growth can sometimes be even accelerated compared to the untreated tumor due to the preference of therapy resistant clones.