Personalised therapeutic approaches to glioblastoma: A systematic review

Abstract

Introduction: Glioblastoma is the most common and malignant primary brain tumour with median survival of 14.6 months. Personalised medicine aims to improve survival by targeting individualised patient characteristics. However, a major limitation has been application of targeted therapies in a non-personalised manner without biomarker enrichment. This has risked therapies being discounted without fair and rigorous evaluation. The objective was therefore to synthesise the current evidence on survival efficacy of personalised therapies in glioblastoma.

Methods: Studies reporting a survival outcome in human adults with supratentorial glioblastoma were eligible. PRISMA guidelines were followed. MEDLINE, Embase, Scopus, Web of Science and the Cochrane Library were searched to 5th May 2022. Clinicaltrials.gov was searched to 25th May 2022. Reference lists were hand-searched. Duplicate title/abstract screening, data extraction and risk of bias assessments were conducted. A quantitative synthesis is presented.

Results: A total of 102 trials were included: 16 were randomised and 41 studied newly diagnosed patients. Of 5,527 included patients, 59.4% were male and mean age was 53.7 years. More than 20 types of personalised therapy were included: targeted molecular therapies were the most studied (33.3%, 34/102), followed by autologous dendritic cell vaccines (32.4%, 33/102) and autologous tumour vaccines (10.8%, 11/102). There was no consistent evidence for survival efficacy of any personalised therapy.

Conclusion: Personalised glioblastoma therapies remain of unproven survival benefit. Evidence is inconsistent with high risk of bias. Nonetheless, encouraging results in some trials provide reason for optimism. Future focus should address target-enriched trials, combination therapies, longitudinal biomarker monitoring and standardised reporting.

Keywords: Glioblastoma; genomics; glioma; personalised therapy; survival.

Figure 1

(A) Glioblastoma presents as a space occupying tumour within the central nervous system, most commonly occurring within the supratentorial region. (B) Personalised medicine aims to tailor therapeutics to individual patients to maximise efficacy: numerous genomic variants have been identified as potential targets in glioblastoma. (C) Molecular targets in malignant glioblastoma cells evolve over time in response to selection pressures, such as exposure to therapies. Non-invasive liquid biopsy holds significant potential as a facilitator of personalised therapy; longitudinal monitoring of tumour biomarkers may permit dynamic optimisation of targeted therapies. (D) The p53 pathway is one of the key cellular signalling pathways in which variants have been identified in malignant glioblastoma cells, including in MDM2 and TP53. The redundancy seen within and between cellular pathways likely contributes to therapeutic failure of single agents. Figure created using BioRender.com.

Figure 2

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of study selection. Following the removal of duplicates, titles and abstracts of 6,738 records were screened and 115 studies were sought for retrieval. An additional 19 studies were identified from ClinicalTrials.gov and by hand searching reference lists of included studies and relevant review articles. Reports for 2 studies were not retrievable and 30 studies that initially appeared to meet inclusion criteria were excluded during full text screening (Appendix B). A total of 93 full text articles and 9 conference abstracts were included in the review.

Figure 3

(A) There were a range of designs of included studies: 7 pilot studies, 2 retrospective analyses, 1 multicohort basket trial, 1 phase 0 trial, 32 phase I trials, 23 phase I/II trials, 31 phase II trials and 5 phase III trials. (B) Newly diagnosed and recurrent disease cohorts were each studied by 40.2% of included trials, 15.7% (16/102) included mixed cohorts and the status of patients was unclear for 3.9% (4/102) of trials. (C) More than 20 distinct types of personalised therapy were studied, including a range of targeted molecular therapies and dendritic cell vaccine approaches. (D) Overall survival efficacy of all personalised therapies; in 62.7% (64/102) of studies there was a survival benefit or personalised therapy appeared beneficial.

Figure 4

Personalised therapy trials were mostly non-randomised and were conducted in more than 20 countries worldwide. The United States conducted the most trials with a total of 42.2% (43/102), followed by Japan with 11.8% (12/102).