Ex-vivo drug screening of surgically resected glioma stem cells to replace murine avatars and provide personalise cancer therapy for glioblastoma patients

Abstract

With diminishing returns and high clinical failure rates from traditional preclinical and animal-based drug discovery strategies, more emphasis is being placed on alternative drug discovery platforms. Ex vivo approaches represent a departure from both more traditional preclinical animal-based models and clinical-based strategies and aim to address intra-tumoural and inter-patient variability at an earlier stage of drug discovery. Additionally, these approaches could also offer precise treatment stratification for patients within a week of tumour resection in order to direct tailored therapy. One tumour group that could significantly benefit from such ex vivo approaches are high-grade gliomas, which exhibit extensive heterogeneity, cellular plasticity and therapy-resistant glioma stem cell (GSC) niches. Historic use of murine-based preclinical models for these tumours has largely failed to generate new therapies, resulting in relatively stagnant and unacceptable survival rates of around 12-15 months post-diagnosis over the last 50 years. The near universal use of DNA damaging chemoradiotherapy after surgical resection within standard-of-care (SoC) therapy regimens provides an opportunity to improve current treatments if we can identify efficient drug combinations in preclinical models that better reflect the complex inter-/intra-tumour heterogeneity, GSC plasticity and inherent DNA damage resistance mechanisms. We have therefore developed and optimised a high-throughput ex vivo drug screening platform; GliExP, which maintains GSC populations using immediately dissociated fresh surgical tissue. As a proof-of-concept for GliExP, we have optimised SoC therapy responses and screened 30+ small molecule therapeutics and preclinical compounds against tumours from 18 different patients, including multi-region spatial heterogeneity sampling from several individual tumours. Our data therefore provides a strong basis to build upon GliExP to incorporate combination-based oncology therapeutics in tandem with SoC therapies as an important preclinical alternative to murine models (reduction and replacement) to triage experimental therapeutics for clinical translation and deliver rapid identification of effective treatment strategies for individual gliomas.

Development and Optimisation of Tumour Treating Fields (TTFields) Delivery within 3D Primary Glioma Stem Cell-like Models of Spatial Heterogeneity

“…Additionally, using these tumour core/edge models’ therapeutic responses to reflect treatment-naïve disease from the typically contrast-enhancing tumour bulk (core), potentially supporting efficacy data for regimens to treat patients where surgical resection is not feasible, as well as in distant brain-invasive margin GSCs (edge), which are more reflective of typical post-surgical residual disease following current standard-of-care treatment, can be assessed. MGMT expression was validated in both OX5 core and edge cell models alongside the presentation of the GSC phenotype through the historically established GSC markers Nestin, CD133 and SOX2 ( Figure 5 A), confirming both the TMZ resistance and intratumoural heterogeneity of the model [ 27 , 28 ]. The expression of these markers persisted despite TTFields treatment, which shows a phenotypic GSC retention throughout the treatment regimen.…”

“…We therefore used our optimised 3D GSC TTFields delivery protocol in a matched MGMT+ (a genetic/cellular marker which confers a poorer clinical TMZ response [ 1 ]) glioblastoma tumour core (typically resected) and edge (typically residual) model [ 27 , 28 ] to test these multi-modal regimens and provide proof-of-concept data around our preclinical methodology ( Figure 5 ). Additionally, using these tumour core/edge models’ therapeutic responses to reflect treatment-naïve disease from the typically contrast-enhancing tumour bulk (core), potentially supporting efficacy data for regimens to treat patients where surgical resection is not feasible, as well as in distant brain-invasive margin GSCs (edge), which are more reflective of typical post-surgical residual disease following current standard-of-care treatment, can be assessed.…”

“…G1 GSC cells were generated in Cambridge within Professor Colin Watts’s laboratory and have been previously used in reports from multiple groups, including our own [ 20 , 25 , 26 ]. The primary glioblastoma OX5 core and invasive edge 3D GSC models were generated from fresh treatment-naïve glioblastoma tissue (based on neurosurgeon guidance) collected from consenting patients undergoing surgery at Sheffield’s Royal Hallamshire Hospital (ethical approval: Yorkshire & The Humber—Leeds East REC 11-YH-0319 (STH15598)), as previously described [ 27 , 28 ]. Only glioblastoma tissue surplus to histological requirements was used and it was resected by the operating surgeon, collected intraoperatively whilst surgery was ongoing, pseudonymised/assigned a unique sample identifier and then rapidly transferred to the research laboratory in a dry, sterile specimen pot (NHS) at room temperature within 10 min of collection from surgery.…”

“…Cultrex ® solution was added to plasticware T-75 (2.5 mL/flask) and Alvetex TM 12-plates (500 μL/well) and incubated at 37 °C for 30 min prior to use to permit polymerisation, before aspiration of any excess coating solution. All cell lines used were kept at early passage (5–12 sub-passages from initial derivation) and cultured in GSC culture medium throughout to maintain the GSC phenotype and remove any non-cancerous cells, as previously described [ 27 , 28 ].…”

“…The primary glioblastoma OX5 core and edge 3D GSC models were generated from fresh, treatment-naïve glioblastoma tissue collected from consenting patients undergoing surgery at Sheffield’s Royal Hallamshire Hospital (Ethical approval: Yorkshire & The Humber—Leeds East REC 11-YH-0319 (STH15598)), as previously described [ 27 , 28 ].…”