TABLE 4.
Limitations and challenges of BC-on-chip models.
| Challenges | Limitations | Putative solutions | References |
|---|---|---|---|
| General challenges | |||
| Fabrication of microfluidic devices is complicated | higher experimental costs, high costs of materials (glass, silicon, polymers, metals, quartz, ceramics), process time can be long and expensive, some physical proprieties of traditional materials are not desirable | use of cheaper and appropriate materials for cell analysis; creation of multifunctional and flexible stickers that can be combined depending on the experimental requirements; chip miniaturization; chip integration with other devices; use of 3D printing technology and 3D printable bioinks | Scott and Ali (2021), Niculescu et al. (2021), Cho et al. (2023) |
| Training the next-generation of highly specialized researchers and technicians to acquire specific skills in nanotechnology | professional background with high degree of interdisciplinarity/multidisciplinarity and transdisciplinarity (engineering, physics, chemistry, molecular and cell biology, stem cell research, clinical science, toxicology, pharmacology, imaging) | adequate types of training in novel technologies, including different omics-based fields (proteomics, interactomics, transcriptomics, metabolomics, etc.), data analysis | Moruzzi et al. (2023) |
| Tissue engineering techniques are complex | technical advances in biomaterials, stem cell engineering, microengineering, microfluidic technologies are necessary for reconstructing the tissue microarchitecture and dynamic spatio-temporal microenvironments, integration of biosensors to monitor tissue functionality and TME quality | recreate specific aspects of tissues for particular applications (drug transportation, monitoring barrier function, inflammatory response, EMT, angiogenesis, cancer cell intravasation and extravasation, study of cell-cell and cell-TME interactions) | Cho et al. (2023) |
| Holistic exploitation of quantitative data; coupling of BC-on-chip and artificial intelligence for future BC management | computational models are necessary to assess accuracy, robustness, reliability, reproducibility, efficiency, and relevance of standardized compounds, assays and biomarkers | exploitation of relationship between BC-on-chip and clinical data for translation of microfluidic platforms to the clinic; integration of multiomics data | Imparato et al. (2022), Moccia and Haase (2021) |
| Reproducing the breast gland on its integrality | each platform is designed to address a specific biological process and may not provide a complete network of biological processes that drive tumor behaviour | co-culture of many cell types of the breast or distant metastatic sites | Moccia and Haase (2021) |
| Sample volume and other features of the sample | tissue biopsies or explants (top-down chips) can be often too large to easily be incorporated into a chip | engineered (bottom-up chip) or natural miniature tissues are preferred, but require adequate natural matrices (Matrigel, collagen, hyaluronic acid, gelatine), or synthetic hydrogels (polyacrylamide, polyethylene glycol-fibrinogen, polylactic acid) | Picollet-D’hahan et al. (2021) |
| cancer-on-chip no recreate tumor at real size (typically > 109 cells) | increase the number and diversity of cells | Imparato et al. (2022) | |
| obtaining fresh tissue is challenging and use of patient sample in on-chip devices is limited | optimization of number, viability, growth of primary cells, limitation of variability in establishment and differentiation of iPSCs | Moccia and Haase (2021) | |
| BC cell lines are not optimal for replicating tumor heterogeneity | lack of suitable BC cell lines for all BC subtypes, such as luminal-B cell line; BC cell lines accumulate mutations during subsequent series of cultivation | use of lesser studied and variable cell lines for all BC subtypes, other than MCF7, MBA-MD-231, and T47D | Moccia and Haase (2021), Li et al. (2023a), Dai et al. (2017) |
| Nanoparticle and therapeutic proteins delivery in the malignant BC cells and TME | study of therapeutic nanoparticles/proteins absorption, distribution, metabolism, elimination and toxicity | construction of platforms to predict nanoparticle behaviour with applications in nanmedicine; use of synthetic cells that synthesize therapeutic proteins on-demand inside tumors, killing malignant BC cells | Krinsky et al. (2018), de Roode et al. (2024) |
| Creation of personalized BC-on-chip for accurate diagnosis, monitoring, and drugs testing | genetic, physiologic, and biometric heterogeneity during BC progression sustains diversity of each individual; extracted biopsy represents a single snapshot at a time and the not represent the actual tumor heterogeneity and provide inadequate information | use of patient-specific iPSCs derived from somatic cells that are reprogrammed to a pluripotent stage, and differentiate into a wide spectrum of cell types | Palasantzas et al. (2023) |
| primary cells can be inaccessible or they van not be isolated from tissues with high purity | patient-derived stem cells become more accessible with minimal invasiveness | Leung et al. (2022) | |
| primary cells often lose rapidly their tissue specific functions and viability in vitro and are not suitable for long-term studies | breast-specific cell type requires specific culture conditions | Leung et al. (2022), Moccia and Haase (2021) | |
| Tumoral ECM is constantly reshaped during BC progression | reconstructing the dynamic TME of the primary BC (BC TME-on-chip); mammary tissue TME possesses specific, complex, dynamic biophysical and biochemical features that are difficult to recapitulate | animal and human-based model systems: BC biopsy or surgical resection, dissection, deconstruction, isolation of patient-specific stromal cells for reconstruction | Li et al. (2023a), Mun et al. (2024b), Liu et al. (2021) |
| human-based model systems: repopulation of decellularized patient-derived scaffolds (PDSs) from fresh-frozen BC biopsies (including biobanked) with BC cell lines (MCF7, MDA-MB-231) | Leiva et al. (2023) | ||
| inclusion of micro-devices able to control nutrients, growth factors, chemokine gradients and other parameters | Moccia and Haase (2021) | ||
| difficulty in isolating specific cell subtype | established biomarkers for each subpopulation are required | Mun et al. (2024b) | |
| most models lack interactions with adipocytes and myoepithelial cells | inclusion of these cell types could enhance the quality of BC TME | Moccia and Haase (2021) | |
| addition of to many cell types affects maintenance of the chip | fabrication of standardized BC-on-chip models based on multiple co-cultured cells | Moccia and Haase (2021) | |
| reproduction of protein network from ECM | new matrix biomaterials with good biocompatibility are need to be synthetized; design novel tumor spheroids with inner biomimetic ECM easily penetrable by vascular network | Li et al. (2023a) | |
| incorporation of various types of immune cells on tumor-on-chip | improving technologies for long-term maintenance of the function of immune cells when immune cells are co-cultured with other cells for development of immunotherapies and tumor vaccines | Li et al. (2023a) | |
| Fabrication of multi-organ-on-chip (some challenges are available also for single organ-on-chip) | BC is a systemic disease, so it is necessary to reproduce the dynamic interaction between BC-distant organs, to study BC cell spreading, seeding and metastasis and evaluate the efficacy and off-target toxicity of anti-BC therapeutics | coupling of potential metastatic niches, incorporation of physical, chemical, and molecular biosensors for multimodal and real-time detection of local factors, and online multiomics analysis | Picollet-D’hahan et al. (2021), Liu et al. (2021) |
| culture media affect cell viability, phenotype, and senescence, so different interconnected organs require a microfluidic circuitry and specific perfusion parameters to reproduce in vivo situation; each breast-specific cell type and other organs specific cells require specific culture conditions | use of media to provide organ specificity, avoid infection risk, inter-sample variability, create patient-specific models; optimization of flow rate, prevention of dilution, assuring sufficient nutrients | Moccia and Haase (2021), Picollet-D’hahan et al. (2021), Mun et al. (2024b) | |
| use of serum leads of lack of reproducibility and standardization which is important for drug screen | use of appropriate media to provide organ specificity | Moccia and Haase (2021) | |
| scaling/extreme miniaturization can cause significant structural reorganization and changes in different organ proportions | use of scaling approaches (proportional, allometric, functional) | Picollet-D’hahan et al. (2021), Wikswo et al. (2013) | |
| engineering an appropriate vascular network | use of endothelial-tumor models instead tubing-based modeling; endothelium can be modelled using HUVECs, organ-specific microvasculature endothelial cells that are not always commercially available, endothelial cells isolated from biopsies or differentiated from iPSCs/mesenchymal stem cells | Picollet-D’hahan et al. (2021), Roudsari and West (2016) | |
| organs/cells must be exposed to biochemical, mechanical, and electrical stimuli for proper development and functioning | perfused medium must be supplemented with biochemical stimuli, such as sex hormones | Picollet-D’hahan et al. (2021) | |
| biophysical factors, such as generation of precise gradient fluid-flow shear stress, affect BC cell phenotype and aggressiveness leading to altered behaviour and increased aggressiveness at the metastatic site | Ortega Quesada et al. (2024) | ||
| real-time and latter analysis for assessment of cross-organ communication can perturb the multi-organ-on-chip microenvironment and affect the detection of low-abundant molecules | minimal or no sample preparation, optimization of the amount of sample collected for molecular analysis | Picollet-D’hahan et al. (2021) | |
| Specific challenges | |||
| Liver-on-chip | induction of pluripotent stem cell-derived hepatocytes for personalized treatments adapted on individual patient’s needs; study of drug efficacy and toxicity | use of iPSC-based liver organoids that can differentiate into multiple hepatic cell types: more mature hepatocyte-like cells, cholangiocytes, stellate cells and Kupffer cells | Olgasi et al. (2020), Dalsbecker et al. (2022) |
| Lymph node-on-chip | simulating ECM and reproduction of the complex internal structure of LNs in vitro | constructing LN-on-chip with more complex structures to study the complex interaction of BC cells with stromal and immune cells | Wang et al. (2024) |
| Bone-on-chip | continuous monitoring of local factors that drive bone metastasis | integrating biosensors can provide non-invasive, continuous monitoring of the experiment progression | Zhang et al. (2023) |
| Brain-on-chip | functionality of brain is very complex and differ from person to person | correct assessment of the ratio of non-neuronal to neuronal cells depending on the brain region (1.2 astrocytes/neuron, 0.46 endothelial cells/neuron, 0.2 microglia/neuron), capillary density and blood flow, differences in structure of grey and white matter, shear stress at the endothelial barrier | Wikswo et al. (2013) |
| Gut microbiome-on-chip | complex fabrication process, small chip capacity, limited lifespan of cells, absence of supporting cells (microvascular endothelium, immune cells, goblet cells, muscle cells, enteric neurons), microvilli, and a mucus layer, possible contamination with impact on anti-BC drug/nutraceutical testing, lack of possibility to incorporate all important families of intestinal flora | use of biocompatible materials, incorporation of sensors, incorporating of iPSCs, developing collagen-based cilia-like structures, co-culturing complex microbial species in diverse media and oxygen condition, co-culture of aerobic host cells and anaerobic microbes | Thomas et al. (2023) |
Abbreviations: ECM, extracellular matrix; HUVECs, human umbilical vein endothelial cells; iPSCs, inducible patient-derived stem cells; PDSs, patient derived scaffolds; TME, tumor microenvironment.