Potential of ex vivo organotypic slice cultures in neuro-oncology

Abstract

Over recent decades, in vitro and in vivo models have significantly advanced brain cancer research; however, each presents distinct challenges for accurately mimicking in situ conditions. In response, organotypic slice cultures have emerged as a promising model recapitulating precisely specific in vivo phenotypes through an ex vivo approach. Ex vivo organotypic brain slice models can integrate biological relevance and patient-specific variability early in drug discovery, thereby aiming for more precise treatment stratification. However, the challenges of obtaining representative fresh brain tissue, ensuring reproducibility, and maintaining essential central nervous system (CNS)-specific conditions reflecting the in situ situation over time have limited the direct application of ex vivo organotypic slice cultures in robust clinical trials. In this review, we explore the benefits and possible limitations of ex vivo organotypic brain slice cultures in neuro-oncological research. Additionally, we share insights from clinical experts in neuro-oncology on how to overcome these current limitations and improve the practical application of organotypic brain slice cultures beyond academic research.

Brain cancer presents a significant challenge in clinical oncology based on their extensive heterogeneity, dynamic cellular plasticity, and limited availability of central nervous system (CNS)-effective therapies that improve prognosis resulting in an unacceptable median overall survival of around 6–15 months post-diagnosis.^1,2 Consequently, there is a pressing need to identify therapeutic vulnerabilities and corresponding pharmacological agents exhibiting robust CNS activity in preclinical models^3,4

In this context, in vitro, in vivo, and ex vivo models have emerged as invaluable assets for brain cancer research throughout the last decades^5,6 (Table 1 and Figure 1).

Limitations of Current Preclinical Models

Nevertheless, each of these preclinical models is associated with distinct obstacles (Table 1).⁷ Patient-derived xenografts, despite their utility, present significant challenges such as prolonged initiation periods, potential animal distress, substantial financial costs, as well as variances to the original tumor.

Similarly, patient-derived cell lines often undergo clonal evolution resulting in both, genetic and phenotypic heterogeneity compared to the original tumor.⁸ While certain patient-derived organoids exhibit the capacity to faithfully replicate the histological and mutational diversity of human cancer, they are predominantly achieved from aggressive, high-grade tumors, thereby potentially introducing a bias in subsequent interpretations.^9,10 Conversely, generating such models from heterogeneous tumor populations poses challenges for reproducibility among intratumor replicates.¹¹

Indeed, although there is a wide range of experimental models available to replicate cancer disease, their usage is hindered by significant economic and animal welfare challenges imposing substantial limitations on the comprehensive validation of initial preclinical findings.

Overcoming Current Limitations in Preclinical Research

To address the limitations of current models, organotypic brain slice cultures have been developed and refined in oncological research over the past few decades. Unlike brain organoids, which are made by differentiating pluripotent stem cells in vitro, ex vivo organotypic brain slice cultures come directly from mouse or human brains via neurosurgical resections.^12,13

As a result, brain organoids generated from induced pluripotent stem cells exhibit greater artificiality and immaturity, limiting their ability to accurately replicate the complexity of the human brain.¹⁴ In contrast, organotypic brain slice cultures maintain the brain’s native architecture, including neuronal networks, synaptic organization, and environmental cells like neurons, astrocytes, microglia, and immune cells.^15–17 Thereby, this model incorporates biological relevance and inter-patient variability, aiding drug discovery and treatment stratification.^18–20 However, only a minority has made it through to direct patient diagnostic or treatment approaches within robust clinical trials.

In this review, we explore the potential of human- or mouse-derived organotypic brain slice cultures in both basic and clinical research, offering a valuable tool for enhancing brain cancer research. Additionally, we share insights from neuro-oncology clinical leaders on the nonacademic applications of ex vivo organotypic brain slice cultures.

Organotypic Brain Slice Cultures as a Successful Example of an Ex Vivo Brain Cancer Model

Culturing cancer cell lines in vitro is essential for investigating cell survival, morphology, functionality, and therapeutic effects.⁷ This method reduces the need for animal experiments and their distress. However, isolated cells do not capture the microenvironment’s complexity due to isolation and lack of intercellular communication. Recently, ex vivo organotypic brain slice cultures simulate more physiologically relevant, in situ-like conditions, revealing their potential in brain cancer research.^21–23

Minimizing Animal Avatars as Preclinical Models for Research Purposes

Conducting in vivo mouse models that reliably replicate clinical phenotypes demands substantial costs, time, and resources. As an illustration, a modest-scale drug discovery investigation using patient-specific xenograft models might necessitate over 25–50 mouse avatars to derive personalized recommendations from a selection of 5 treatment options for a single patient (5–10 mice/treatment; Figure 2A)²⁰ As a result, constraints related to time, expense, and ethical considerations enforce limitations on the number of treatments feasible to test using patient-specific xenograft avatars for a single patient.¹⁸

In contrast, ex vivo organotypic slice culture models offer the advantage of cutting a single tumor bulk into multiple slices depending on the sample’s size and quality. This approach allows for the simultaneous evaluation of multiple anticancer drugs and the straightforward application of the main principles of animal research within a single avatar^13,24 (Figure 2B). In addition, in the case of patient-derived tumors, this approach enables the complete omission of animal models for investigational processes (Figure 3C).

In detail, mouse brains with established brain tumors, developed 5–7 weeks post intracardiac or intracranial injection of human primary or secondary brain cancer cell lines, are resected and embedded in 4% low-melting-point agarose. The brains are then sectioned into 250-μm slices using a vibratome. These tissue slices are subsequently cultured on membranes in appropriate culture media and incubated at 37 °C for up to 1 week¹⁵ (Figure 3A).

This preparation method allows for the evaluation of advanced brain tumors, facilitating pharmacological and biological investigations with clinically relevant established tumors. In contrast, to analyze the initial steps of normal brain colonization, human or mouse cancer cell lines are placed on tumor-free brain slices obtained from mice and incubated under the same conditions as described above²⁰ (Figure 3B). Periodic bioluminescence or confocal imaging can be conducted to accurately monitor tumor infiltration and growth.^15,25

In studies including human brain tumors, human brain samples obtained during routine neurosurgical resections can be immediately sectioned, cultured, and incubated as previously described, without any use of mouse avatars¹⁵(Figure 3C).

Additionally, slices can be cultured in the presence of various therapeutic agents to identify potent antitumor drugs or radiated to assess the response to radiotherapy. At the end of the experiment, fixation with 4% paraformaldehyde overnight at 4 °C is performed before doing further analysis.¹⁵

Mirroring In Vivo Architectural, Neuronal, and Genetic Features in an Ex Vivo Model

The advantages of ex vivo organotypic models in cancer research are beyond their economic conditions; they also present many complex replications of in vivo conditions.

Maintenance of cytoarchitecture.—

Ex vivo organotypic brain slice cultures excel in preserving the structural coherence and synaptic functionalities of human brain tissue for up to 1 week. In particular, immunohistochemical studies on cultured glioblastoma (GBM) slices have shown that the maintenance of cytoarchitecture in human cortical explants does not require serum supplementation.^16,26 Although a temporary activation of astrocytes and microglia is noted in freshly cut slices- primarily associated with mechanical stress from resection and sectioning- this activation decreases within the first 24 hours.²⁷ Subsequently, no additional variances are observed throughout the culture period, even in the absence of external growth factors.¹⁶

Preservation of neuronal electrical activity.—

Furthermore, the progression of high-grade gliomas is significantly influenced by neuronal electrical stimulation.^28,29 While several models failed to capture the complex electrical interplay, ex vivo organotypic brain slice cultures have demonstrated the ability to reflect this network activity. Importantly, during repetitive recordings of extracellular electrical activity, stable neuronal action potential could be observed in patient-derived organotypic brain cultures.^16,17,30 This highlights the critical role of ex vivo organotypic brain slice models in mimicking the electrophysiological conditions crucial for studying the progression of brain cancers like high-grade gliomas.

Reflecting genetic signatures.—

Given the growing significance of personalized medicine in cancer therapy, maintaining the gene expression profiles of cultured brain cancer slices over time is critical. Vaira et al. conducted a study to assess the genetic and phenotypic stability by culturing organotypic tumor slices from 13 different primary entities alongside normal organ tissue for 5 days.²¹ They confirmed the preservation of overall architecture, including epithelial structures and their spatial relationships with stroma, in both normal and tumor samples throughout the 5-day culture period. Furthermore, they found that cell viability, pathway activity, and global gene expression profiles were also preserved ex vivo for up to 5 days.

In line with these results, Ravi, Joseph, and collaborators analyzed the top 500 up-and-down-regulated genes with RNA sequencing analysis and compared these between freshly sectioned human GBM slices and human GBM slices cultured for 1 and 2 weeks.¹⁶

These findings underscore that ex vivo organotypic brain slice cultures offer the ultimate opportunity to replicate the complex in vivo morphology, cellular structure, electrical neuronal interplay, and gene expression signatures from a few days up to 2 weeks in culture, which presents the closest representation to in vivo conditions to date.

Incorporation of the Tumor Microenvironment Dynamics in an Ex Vivo Model

Exploring the initial steps of normal brain colonization.—

The brain microenvironment serves as a critical element in the initial phases of brain metastasis (BM) colonization influencing the survival, expansion, and proliferation of metastatic-initiating cells.³¹ Remarkably, despite the daily spreading of thousands of cells from the primary tumor, only around 0.02% manage to endure in circulation and establish micro-metastases in highly aggressive in vivo models.³² These steps necessitate the ability to evade organ-specific defense mechanisms within the adjacent brain parenchyma and the adaptation to supportive niches. Consequently, it is of utmost importance to focus on studying this subset of cells capable of withstanding this glial defense, as they predominantly dictate patient outcomes.³³

In this context, ex vivo organotypic brain slice cultures are valuable models for replicating the phenotypic and functional characteristics of early BM. These models have been used to study mechanotransduction signaling from cancer cells spreading on capillaries. To delve deeper into this cascade, metastatic cells were plated on mouse brain slices and observed with time-lapse confocal imaging showing cell infiltration, migration toward capillaries, and L1CAM-mediated spreading.³⁴ This research highlights the model’s suitability for understanding post-extravasation steps and L1CAM-associated metastatic recurrences.

Another study identified plasmin as a key defense factor against metastatic infiltration using ex vivo organotypic models.³⁵ In detail, H2030-BrM3 cells were placed on cultured mouse brain slices and showed migration, proliferation, and interaction with blood capillaries, unlike parental H2030 cells, which underwent apoptosis. Co-culturing with astrocytes and microglia led to the conversion of plasminogen into plasmin, triggering apoptosis in parental H2030 cells. In contrast, plasmin inhibitors increased their survival. These findings highlight plasmin’s role in targeting cancer cells in the brain and the ability of metastatic cells to evade these effects, providing new insights for clinical research.

In-depth analysis of the invasive fronts in brain metastasis development.—

However, the brain microenvironment not only serves as an essential factor in the initial colonization of brain metastatic cells but also has a dominant role during the presence of established metastases. While micro-metastases often still lack a clear border zone between the metastatic and parenchymal cells, macro-metastases in contrast disrupt later the intervening organ parenchyma and establish their own tumor microenvironment.³¹ During this process the macro-metastasis/organ parenchyma interface (MMPI) develops, defining the region where cells from the macro-metastasis and its surrounding microenvironment directly engage with the adjacent organ parenchyma.³⁶ However, the intricate development of the MMPI has not been adequately demonstrated for long.

In this context, ex vivo organotypic brain slice cultures effectively replicate the phenotypic and functional features of MMPI. In detail, a co-culture model with mouse brain slices and a 3D cancer cell matrix in matrigel was developed to mimic MMPI.³⁷ This setup allows for continuous monitoring of metastatic colonization, revealing morphological changes and cell interactions through fluorescence and bright field microscopy. Indeed, studies using ex vivo brain slice cancer cell co-cultures found that modulating glial defenses with CXCR4 or WNT inhibitors, or bisphosphonates, reduces cancer cell infiltration into the brain. This suggests that targeting glial defenses at the MMPI could offer a new therapeutic strategy to improve treatment outcomes.^38,39

In addition, diverse histological growth patterns of the MMPI were detected across different human cell lines in organotypic co-culture models.³⁶ A surgical biopsy study analyzed BM tissues from patients with different primary tumors, revealing that BM from lung, breast, melanoma, and colorectal cancer showed infiltrative patterns, while renal cell cancer had displacing patterns.⁴⁰ Importantly, infiltrative patterns were linked to a 3-fold shorter overall survival. The variability in growth patterns suggests underlying molecular mechanisms and potential therapeutic vulnerabilities, warranting further investigation in ex vivo models.

Targeting the tumor microenvironment in brain metastasis as an emerging therapeutic strategy.—

The contribution of the tumor-associated microenvironment is notably absent in in vitro assays. However, the ability to assess the intricate interplay between tumor-associated brain microenvironment and cancer cells is crucial for therapeutic response in brain tumors.⁴¹ Indeed, the tumor microenvironment in cancer encompasses not only tumor cells but also stromal, immune, and vascular cells, along with extracellular matrix components. Within this environment, regulatory T cells, tumor-associated macrophages, and myeloid-derived suppressor cells lead through intricate pathways to the immunosuppressive and exhausting nature observed in brain cancer.⁴²

So far, brain cancer research has mainly relied on genetically modified, cell line xenograft and patient-derived xenograft mouse models. However, they exhibit not just altered immune responses but also significant interspecies differences.^43,44 Indeed, these models fail to capture the full complexity of interactions among tumor cells, stromal and immune components, as well as therapeutic agents (Table 1). Tumor cell lines alone do not adequately represent the antigen surface expression patterns of intracerebral tumor tissue, revealing altered immune activities and significant interspecies differences.^34,35 Moreover, patient-derived organoid models need additional steps to accurately reflect the tumor microenvironment, requiring specific co-culturing of blood-derived immune cells. Importantly, it remains unclear whether co-culture methods might alter the composition of the tumor microenvironment in brain cancer organoid models, potentially introducing another bias.^42,45

The following sections discuss studies that examine the composition of the tumor microenvironment in organotypic brain slice cultures, which have subsequently led to clinical trials in neuro-oncology.

• - Reflection of the tumor microenvironment in organotypic brain metastasis slice cultures.

The central nervous tumor microenvironment consists mainly of astrocytes (40%) and microglial cells (10%–15%). Consequently, the reactive changes of these cells play an essential role in the progression, infiltration, and angiogenesis of brain tumors.

In line, Priego and colleagues used organotypic brain slice cultures to evaluate drugs targeting the pro-metastatic microenvironment, focusing on pSTAT3 + reactive astrocytes rather than just cancer cells.⁴⁶ They employed a genetically engineered mouse model with tamoxifen-inducible STAT3 knockout and shRNA against STAT3. By comparing gene silencing in cancer cells and astrocytes with inhibitors targeting both, they discovered a novel pro-metastatic role for pSTAT3 + reactive astrocytes in BM. These findings led to a clinical study with the STAT3 inhibitor silibinin, showing promising results in brain metastatic lung cancer patients.⁴⁷ A phase II trial is now assessing silibinin’s efficacy in preventing CNS recurrence after BM resection (NCT05689619).

Another recent study by Monteiro, Miarka, and colleagues highlights the importance of incorporating the tumor microenvironment in researching radio-resistance biomarkers in BM. The study identified the S100A9–RAGE–NF-κB–JunB pathway as a key mediator in radio-resistance.⁴⁸ They found that S100A9 expression in brain metastatic cells promotes resistance to radiotherapy via NF-κB activation. Targeting S100A9 by a blood-brain barrier (BBB) permeable inhibitor of its receptor, RAGE (receptor for advanced glycation end products), sensitized BM to irradiation in patient-derived brain slice cultures. In line, a phase I trial is currently investigating the safety and efficacy of azeliragon, a RAGE inhibitor, combined with standard radio-chemotherapy for GBM (NCT05635734). The initiation of another phase II trial, which explores the combination of azeliragon with brain radiotherapy for BM, is currently planned.⁴⁹

• - Reflection of the tumor microenvironment in organotypic glioblastoma slice cultures.

Given that GBM exclusively occurs within the brain, the CNS tumor microenvironment is crucial for GBM maintenance and proliferation. GBM features a highly immunosuppressive microenvironment, which hinders the effectiveness of immune-based vaccines and checkpoint inhibitors.^50,51 Therefore, optimizing the interaction between the tumor and its microenvironment in preclinical models is essential, potentially even more so than for BM.

A recent study investigated the expression of antigen markers within the cellular tumor microenvironment of human GBM using 2 3D ex vivo patient-derived models: organotypic brain slice cultures and patient-derived organoids.⁴² The study found that patient-derived organoids failed to maintain the cellular inflammatory characteristics observed in the form of reduced or absent markers for cytotoxic or regulatory T cells, or the absence of specific macrophage markers, while organotypic brain slice models sustained various types of tumor microenvironment cells for up to 9 days. In detail, there was a decrease in CD45+, CD7+, and CD4 + T cells in organotypic brain slices, though they remained present during the 9-day observation period. In contrast, markers of macrophages (CD68, CD163, and CCR2) maintained expression levels similar to primary tissue. Regulatory T cells were absent by the end of the culture period, as indicated by the lack of CD25 and FOXP3 expression on days 8 and 9. Initial levels of PD1, TIM3, and LAG3 were low and decreased further during culturing. Indeed, a notable limitation of the organotypic culture slices was their inability to be sustained beyond 9 days, which restricts their use for longer-term immunotherapy studies.

In contrast, co-culturing patient-derived organoids with peripheral blood mononuclear cells from healthy donors preserved cellular tumor microenvironment patterns for up to 3 weeks, enabling longer experimental setups. Nevertheless, the co-culture organoid model exhibited a lower representation of monocytes and macrophages compared to the organotypic models. In addition, the study did not investigate whether co-culture methods might change the composition of the tumor microenvironment in brain cancer organoid models, potentially leading to an additional bias.

Another preclinical study analyzed single-cell gene expression from over 45 000 immune cells in 12 tumor samples and identified HMOX1 + myeloid cells that release Interleukin-10, causing T-cell exhaustion and contributing to GBM’s immunosuppressive environment.⁵² Human neocortical GBM tissue slice cultures, inoculated with patient-derived T cells, mirrored this T-cell dysfunction. Inhibiting the JAK/STAT pathway with ruxolitinib restored T-cell function, demonstrating proof of principle. Based on these results, a phase I study of ruxolitinib combined with standard care in 60 grade III glioma and GBM patients showed promising efficacy and tolerability (NCT03514069), with a phase II trial planned.⁵³

Revolutionized Drug Screening in Primary and Secondary Brain Tumors

Despite the evident need, clinical trial options for primary and metastatic CNS cancers are scarce, resulting in a detrimental loop that slows down the progress of new therapeutic possibilities for these patients with high unmet clinical needs.⁵⁴ One major challenge in clinical trials is the lack of CNS efficacy, as numerous therapies that initially showed promising efficacy in preclinical models have later failed in patients.⁵⁵ Indeed, murine models of GBM failed to truly mimic the human conditions in GBM, whereas the use of human GBM cell lines in animals can cause cross-species–specific activities that can further lead to wrong conclusions on response to treatment.⁵⁶

Drug Screening in Primary Brain Tumors

To demonstrate the utility of ex vivo organotypic brain slice culture for drug screening in GBM, the response of temozolomide (TMZ), a well-established chemotherapeutic drug in the treatment of GBM, on patient-derived glioma cell lines cultured on healthy human brain slices was tested.¹⁶ Analysis of tumor size and quantification revealed reduced or arrested tumor proliferation in TMZ-sensitive (=methylated MGMT-glioma) cell lines, while the TMZ-resistant (=MGMT-unmethylated glioma) cell lines showed only minimal response to TMZ therapy. These findings are in line with observations on the divergent efficacy of TMZ in methylated and unmethylated GBM within clinical trials,⁵⁷ thus validating the utility of ex vivo organotypic brain culture techniques for investigating the effects of therapeutic agents on tumor growth. Consequently, drug screening initiatives using ex vivo organotypic brain slice cultures have been expanded to explore the efficacy of anticancer agents with limited or unknown activity within the CNS.^58,59

In line, Mann and collaborators have recently published a comprehensive platform based on ex vivo organotypic brain slice cultures, complemented by a multi-parametric algorithm.⁶⁰ This innovative framework facilitates rapid engraftment, treatment, and analysis of uncultured patient brain tumor tissue, as well as patient-derived cell lines spanning the spectrum over high- and low-grade adult and pediatric brain tumors. The algorithm precisely calculates dose–response relationships for both tumor eradication and off-target toxicity for each drug, thereby providing comprehensive drug sensitivity scores predicated on therapeutic efficacy. Notably, this approach allows for the normalization of response profiles across a set of FDA-approved and investigational agents. Furthermore, the summarized patient tumor scores after ex vivo treatment exhibit compelling correlations with clinical outcomes, underscoring the promise of the organotypic brain slice culture platform in precise functional testing to form clinical decision-making. Currently, a clinical trial is planned to explore the utility of organotypic brain slice cultures as living tissue substrates in the context of primary brain tumors.⁶⁰

Drug Screening in Secondary Brain Tumors

In BM, the METPlatform presents such a pioneering medium-throughput drug screening platform based on ex vivo organotypic brain slice cultures, enabling the evaluation of inhibitors against BM growing in situ.²⁴ In detail, the METPlatform was established to investigate the vulnerabilities of macro-metastases in order to therapeutically target clinically relevant stages of BM development. This involved the comprehensive analysis of a chemical library consisting of 114 FDA-approved drugs or agents currently undergoing clinical trials in organotypic cultures. Through this approach, the METPlatform could identify heat shock protein 90 (HSP90) inhibitors as potential targets to enhance the susceptibility of BM. Subsequently, as a proof of principle, the efficacy of the HSP90 inhibitor DEBIO-0932 was validated in both, experimental and human specimens, where biomarkers of response to this drug identified a group of patients with poor prognosis. In general, human samples could be easily integrated into this drug screening platform, facilitating a more comprehensive validation closer to the clinical scenario.

During drug screening processes, the assessment of toxicities is an essential aspect. Importantly, organotypic brain slice cultures serve as a valuable resource for investigating the toxicities of therapeutic compounds.^24,61 For example, using specific markers for neurons and endothelial cells enabled the exclusion of widespread nonselective cytotoxicity within this organ.

Endpoint Analysis in Drug Screening Processes

The endpoint analysis presents an important part of drug screening within ex vivo organotypic brain slice cultures. Currently, 2 options can be used to evaluate markers to assess drug response.

One option is to homogenize the organotypic models. The homogenization allows for the isolation and measurement of proteins, DNA, RNA, or metabolites using mass spectrometry or genomic, transcriptomic, and metabolomic profiling.^24,62,63

On the contrary, keeping the organotypic cultures intact, spatial biomarker analysis presents the advantage of investigating the relationship between drug responses with specific tumor characteristics like heterogeneity within the same sample.⁶²

Recent advancements in multispectral imaging technology now enable the investigation of not only tumor cells but also the differentiation of various cell types within the tumor microenvironment such as immune-related cells or astrocytes.⁴⁶ Indeed, the accurate identification of immune-cell subsets often necessitates the use of multiple markers. Therefore, conventional immunohistochemistry techniques, which typically stain only one protein at a time, are not suitable for this complex process.^64–67

An alternative approach involves performing multiplex fluorescent immunohistochemistry enabling the simultaneous visualization of multiple proteins by confocal imaging with enhanced antibody penetration and requires less tissue compared to traditional immunohistochemistry.⁶⁸ The technique allows for the evaluation of the specific therapeutic effect of an investigational agent on specific cell types and further stratifies outcomes in both, tumor and non-tumor regions by using antibodies against specific markers such as Ki67 (cell proliferation), glial fibrillary acid protein (GFAP) (reactive astrocytes), Olig2 (oligodendrocytes), or Iba1 (microglia).¹⁵

Thereby, this analysis enables the differentiation of responders and non-responders to a given therapy. Additionally, considering the possibility of performing immune-cell profiling of different cell types, multiplex fluorescent immunohistochemistry further enables the identification of the localization of immune-cell subsets and their evolutions during drug response.⁴⁶

Obstacles and Limitations of Ex Vivo Organotypic Brain Slice Cultures

Despite their numerous benefits compared to other preclinical models, various challenges and limitations at different stages of organotypic brain slice culturing processes have to be considered before initiating preclinical or clinical studies.

Missing Standardized Protocols for Resection, Processing, and Culturing

At the moment, we lack standardized protocols for uniformly collecting and processing human or mouse tissue to preserve its structural integrity. Adult or aged animal brain tissue is particularly prone to damage during slicing, requiring expertise to reduce this issue.⁶⁹ The time between tissue resection and slicing is also critical.⁷⁰ A recent study found a 60% reduction in apoptotic cells when slices were cut within 5 minutes of decapitation compared to after 5 minutes.⁷¹ However, tumor necrosis might occur in situ and not due to suboptimal culture conditions.²²

Therefore, standardized protocols are essential for reproducibility and comparability between research institutes. In addition, efficient collaboration between clinical and preclinical teams is also crucial for the timely processing and availability of human tissue samples.

Maintenance and Stability for Long-Term Experiments

Long-term culturing of human brain slices under organotypic conditions is extremely beneficial, as it greatly prolongs the duration of experiments. However, when attempting to culture slices from adult animals or humans, maintaining viability, neuronal activity, and structural integrity usually does not exceed a few days to 2 weeks posing limitations of organotypic brain slice cultures in terms of long-term experiments.

Neuronal networks over extended time.—

Maintaining the structural and electrophysiological integrity of organotypic brain slice cultures over extended ex vivo periods poses a significant challenge for studying neurons and neuronal networks.⁷² Addressing this issue, several studies have explored potential solutions. In 2014, Eugene et al. introduced a protocol detailing the long-term maintenance of organotypic temporal lobe slice cultures from adult epileptic patients, achieving up to 4 weeks using a newly formulated culture medium.⁷³ Additionally, Bjorefeldt demonstrated the stimulating effects of human cerebrospinal fluid on adult rodent organotypic brain slice cultures.⁷⁴ Indeed, the beneficial effect of human cerebrospinal fluid was also evident in human organotypic cortical slices, preserving neocortical neuronal viability and robust single-cell and network electrophysiological function for 2–3 weeks.^72,75 However, methods enabling cultures to extend beyond 4 weeks have yet to be established.

The thickness of brain tissue slice.—

Another important factor is the thickness of slices, ideally between 200 and 400 μm, mainly prepared by using a vibratome.^13,15,72 It’s crucial that slices do not become significantly thinner at the time of cutting as recent studies indicated a significant reduction in thickness of 20%–30% within hours of slicing compared to the initial setting on the chopper.⁷⁶ This thinning tends to occur more noticeably after about one week, according to older studies, which may further impair the viability and structural integrity of slices in long-term culturing.⁷⁷ Therefore, accurately measuring slice thickness may be an important step in specific experiments, in particular, when assessing and preparing brain slice cultures for long-term experiments.^77,78

On the contrary, thicker slices may be associated with central necrosis caused by ischemia, resulting from a disrupted circulatory system and subsequent limited diffusion of nutrients and oxygen supply.⁷⁹

To address challenges with thicker organotypic brain slices (700 μm), a recent study used an interstitial microfluidic perfusion method with a micro perfusion chamber.⁸⁰ This method provided a continuous flow of the oxygenated medium, removing waste and extending viability. It preserved the functional activity and cytoarchitecture of thick brain slices for up to 5 days, offering a viable alternative when traditional methods are not feasible.

Lack of natural vascularization.—

Indeed, the lack of natural vascularization in organotypic brain slice cultures presents a challenge not only for thick slices but also limits efficient oxygen and nutrient diffusion within the central regions of all slices. This restriction could lead to a necrotic center that produces artifacts when testing potential therapeutic agents.⁷⁹

Given the lack of a natural circulatory system facilitating nutrient delivery and waste removal, it can be necessary to implement perfusion systems with a turnover time of a maximum of 2 days for specific experiments, especially for experiments over longer periods.^71,80

Here, a study investigated a novel perfusion system consisting of silicone tubes, which were connected to a petri dish with organotypic mouse brain stem slice cultures.⁷¹ Through this constant circulation of culture medium, neuronal survival was prolonged up to 2 weeks ex vivo with a peak in tissue and cellular quality around 7 days in vitro. In total, an improvement in cell survival for up to 14 days was observed with this perfusion system.

However, the absence of perfusion systems capable of maintaining the stability and viability of organotypic cultures over several weeks or months remains a significant limitation. This is particularly problematic when the aim is to analyze the long-term response of drugs in vitro.

Besides, even without intact capillaries and blood flow, capillary-like structures can survive and release molecules that impact neighboring cells, such as nerve fibers.⁸¹ Recent research shows that brain vessels in organotypic brain slices can regenerate with exogenous stimuli like fibroblast growth factor-2 or vascular endothelial growth factors.⁷⁹ Fundamental capillary components can persist for up to 4 weeks without blood circulation, suggesting that blood perfusion and pressure gradients may not be absolutely essential for endothelial cell survival or capillary regeneration.⁸¹

Integration of the BBB

Nonetheless, ex vivo organotypic brain slice cultures possess limitations as an experimental platform for drug screening as they are not able to provide insights into the penetration of potential emerging therapeutic agents through the BBB.

The BBB functions as a highly specialized regulator, managing the transport of cellular components, molecules, and ions between the systemic blood circulation and the cerebral parenchyma to maintain brain homeostasis. Historically, the BBB has posed a serious obstacle in the intracranial response of therapeutics, as the vast majority of macro-molecule agents and also most of the micro-molecule agents were not able to cross the BBB.⁸² This highlights the necessity of investigating the permeability of newly developed therapeutic agents through the BBB in preclinical 3D models to enable more precise and clinically orientated therapeutic stratifications.

However, there is a current lack of established ex vivo organotypic brain culture models that integrate both, brain tissue and a functional BBB.⁸³ Therefore, so far, various methods of concurrent BBB modeling have emerged from the early 2D static models to the current 3D dynamic microfluidic chips and BBB-spheroids, each presenting distinct advantages and limitations.^84–86

These models exhibit significant potential to improve drug discovery, particularly when associated with computational methodologies. Additionally, the current dynamic evolution of computational programs provides novel predictive tools for evaluating the BBB permeability characteristics of pharmacological compounds^87–89

Obstacles in Endpoint Analysis

Another notable limitation of patient-derived organotypic brain slice cultures in brain cancer trials is the challenge associated with endpoint analysis. While tumor progression can be readily monitored over time in cultured brain metastatic slices from mice using bioluminescence imaging with D-luciferin,^13,15 evaluating intact and developing tumor tissue in patient-derived models and unmodified primary cultures is considerably more complex, potentially more expensive, and necessitates advanced techniques as described before.

Therefore, conducting studies with patient-derived organotypic cultures requires not only expertise in processing and culturing patient-derived material, but also competence in endpoint analysis, including genomic, transcriptomic, and metabolic evaluations, as well as expertise in optical microscopy imaging.

Furthermore, research institutions must have access to and availability of these advanced technologies to include patient-derived organotypic culturing methods in larger preclinical and clinical studies. Collaborating with national and international research institutions with expertise in these specific fields might be helpful to overcome the potential limited access to these advanced techniques.

Ex Vivo Organotypic Brain Slice Cultures in Clinical Research: Challenges to Overcome

Despite the success of organotypic brain slice cultures in preclinical studies, they have seen limited clinical trials in neuro-oncology. To understand the obstacles, we consulted neuro-oncology experts for their views on integrating ex vivo organotypic brain slice cultures into clinical research.

Availability of Fresh Human Brain Tissue

Access to fresh tissue is a major barrier, with many institutions lacking clinical collaborators or brain tissue banks. Indeed, the quality and quantity of samples are also concerns for reliable organotypic slice cultures. Therefore, standardized protocols and better collaborations between clinical and preclinical teams are needed to improve the accessibility and integrity of human tissue samples in neuro-oncological research.

In this context, RENACER in Spain is a leading example, efficiently collecting and processing fresh human brain tissue from 12 hospitals, which is then stored at the Centro Nacional de Investigaciones Oncológicas (CNIO) and shared with national and international research institutions.¹⁹ This model of collaboration could be replicated in other countries to enhance fresh tissue availability for research.

Reproducibility and Stability

Reproducibility is a critical issue highlighted by experts. They stress these models to better optimize for large-scale drug screening to ensure reliability in clinical research. Additionally, maintaining these cultures over time is challenging due to their limited lifespan, which can shorten the period for experimental duration and data collection. To overcome the challenges posed by differing approaches, again the establishment of standardized protocols is essential. Effective collaboration between clinical and preclinical teams is vital. The implementation of uniform protocols, as advocated by professional societies, would not only address these challenges but also significantly improve interlaboratory reproducibility, which is currently hindered by methodological heterogeneity.

Diagnostic Medical Device Regulations

In 2022, the European Commission published the In Vitro Diagnostic Medical Devices Regulation (IVDR; EU) 2017/746, replacing previous directives with uniform regulations across Europe.⁹⁰ The IVDR imposes stricter documentation and review requirements and introduces a risk-based classification system from A (lowest risk) to D (highest risk), placing more IVDs in higher-risk categories.⁹¹ For example, cancer tests are automatically classified as risk class C (the second highest), regardless of the specific use or expertise. This increased stringency may limit the use of experimental 3D technologies in clinical trials and routine decision-making.

Future Steps to Enhance the Inclusion in Clinical Trials

To enhance the attractiveness and viability of ex vivo organotypic brain slice cultures as a foundation for clinical trials, the following points are essential to adapt or change in the future.

Support for prospective tissue sampling and the availability of standardized procedures (SOPs) is crucial for reproducibility. Indeed, publishing and harmonizing detailed protocols, and facilitating technology transfer between research centers, are key to improving tissue culturing success and achieving faster, robust results. Additionally, establishing and publishing SOPs for generating, plating, and maintaining ex vivo patient-derived brain slice cultures by organizations like the European Association of Neuro-Oncology would further help and accelerate the IVDR approvals.

These SOPs must detail necessary resources, personnel, and equipment for clinical trials based on published protocols of the last years which have demonstrated efficacy in brain cancer research.^13,15,16,72

In the future, establishing national strategies could lead to a global initiative enhancing the feasibility and benefits of clinical trials in brain tumors. Collaborations, national and international tissue banks, joint ethics agreements, and international consortia for sharing protocols and materials within the neuro-oncology community are essential for integrating organotypic brain slice cultures into trial planning.

Conclusions

In conclusion, ex vivo organotypic brain slice cultures are a valuable model for studying cancer cell interactions with brain components during disease progression. They accurately replicate early and advanced stages of brain colonization, thereby providing a reliable tool for investigating tumor initiation and subsequent progression steps. In addition, they also facilitate testing drugs not previously used for CNS cancer, allowing prioritization for in vivo validation. However, limitations including the inability to assess drug penetration through the BBB, lack of natural vascularization, and long-term stability have to be considered. Recent advances like 3D microfluidic chips, perfusion systems, specific culture mediums, and computational methods have reduced these issues.

Another major obstacle to using ex vivo organotypic brain slice cultures in clinical research is the lack of standardized protocols. This review urges the neuro-oncological community to create and share SOPs for generating, plating, and maintaining patient-derived organotypic brain slice cultures for diagnostics and treatment decisions. Led by a consortium of specialists, this initiative will improve technology transfer between research centers, enhancing the efficiency, consistency, reproducibility, and clinical applicability of organotypic brain slice cultures in neuro-oncology.

To conclude, ex vivo organotypic brain slice cultures are a pioneering model for brain cancer research, replicating in vivo cancer cell characteristics and the tumor microenvironment. Additionally, they enable simultaneous and rapid assessment of numerous experimental conditions in an economical and accessible manner, thereby, reducing animal use which supports the 3R principle (replacement, reduction, and refinement). To maximize their potential, the neuro-oncological community should develop standardized protocols and establish tissue transfer networks, establishing organotypic brain slice cultures as a key model for brain cancer research and treatment decisions.

Acknowledgment

The authors would like to thank Cary Anders, Petra Hamerlik, Michael Platten, and Frank Winkler for their valuable participation in an online questionnaire regarding their opinions on the use of organotypic cultures in clinical research.

Conflict of interest statement

M.V. declares receiving research funds from AstraZeneca. The other co-authors do not have any conflicts of interest.

Funding

This work was supported by AECC (PRYCO234528VALI; M.V.), La Caixa (HR23-00051; M.V.), ERANET Cofund TRANSCAN-3 (EC co-funded JTC 2021 (M.V.) with funds from Instituto de Salud Carlos III/ NextGenerationEU/ PRTR (AC20/00114) and FC AECC (TRNSC213878VALI), ERC CoG (ID: 864759; M.V.) as well as a grant (J4724) from the “City of Vienna Fund for innovative interdisciplinary Cancer Research” (A.S). The authors are grateful for the large coworkers in RENACER (https://renacerbrainmet.org/) as well as the patients who have donated their samples.

References