'Take 5': Precision by Design

This week’s Take 5 selection reflects a field in which organoid and microphysiological platforms are being put to increasingly precise and consequential use as tools for patient stratification, drug discovery, immune modelling, and clinical decision support across oncology.   Across the five articles, a coherent picture is emerging, highlighting the next frontier is not building better models but deploying smarter ones.

Translational progress in oncology is accelerating, though disparities remain.

Skowronska et al. provide the clearest system-level view, mapping 139 registered patient-derived organoid (PDO) clinical trials from 2023 to 2025.  Oncology accounts for 81% of activity, with breast, colorectal, pancreatic, and gynaecological cancers overrepresented relative to the broader trial landscape, reflecting where organoid derivation pipelines are most mature.  Organoid-guided clinical decision-making now accounts for 20% of registered trials, all within oncology.  However, outside oncology, the picture is strikingly different.  Despite substantial preclinical evidence in IBD, COPD, and metabolic disease, non-oncology indications contribute to just 19% of trials and none include organoid-guided decision-making.  The authors frame this as three overlapping waves of translation, with the first wave (oncology) already delivering measurable patient benefit, the second (non-oncology) scientifically advanced but clinically stalled, and the third (immune-competent systems, organ-on-chip, AI-enabled platforms) still preclinical.  Geographically, trial activity is concentrated in China and Italy, underscoring uneven global readiness.

Computational precision is transforming how we screen.

Gillman et al.  demonstrate a different kind of translational leverage.  Rather than screening hundreds of drugs per patient, their TARGET-SL bioinformatics framework uses tumour sequencing data to predict cancer-essential genes and corresponding small-molecule inhibitors via synthetic lethality modelling, then validates predictions in HCC cell lines and patient-derived liver cancer organoids.  The results are striking.   Hit rates of 37.  5% in cell lines and 25% in PDOs, compared to under 5% in conventional high-throughput ex vivo screens.  Standout compounds include AZD5582 (targeting a BIRC3 gain-of-function mutation in cholangiocarcinoma organoids) and Mithramycin A (targeting SP3), each demonstrating target-specific toxicity while sparing non-tumour hepatocyte controls.  The authors are explicit that this approach is tumour-type agnostic, positioning it as a broadly applicable workflow for reducing the cost and scale of ex vivo drug discovery while improving the signal-to-noise ratio of precision oncology screens.

The tumour microenvironment is the new frontier for organoid fidelity.

Two oncology-focused reviews this week converge on a common message: epithelial-only organoid models are no longer sufficient for the questions the field is asking.  Sani et al. report that across 58 TNBC-specific PDTO studies published from 2015 to 2025, 3D PDTOs remain strikingly underutilised for the immunotherapy bottlenecks that matter most.  No published studies met inclusion criteria for organoid-based T cell exhaustion modelling, only one addressed tumour heterogeneity in the immunotherapy context, and just two investigated macrophage dysregulation.  The dominant application, drug response screening (37% of studies), uses submerged BME culture as the near-universal platform, while the ALI and microfluidic systems best suited for immune-relevant co-culture remain underrepresented.  Ma et al., reviewing gastric cancer PDCOs, frame the problem similarly.  Epithelial monocultures frequently fail to predict clinical drug responses because CAFs, immune components, and the intragastric microbiota all modulate drug sensitivity in ways that monocultures cannot capture.  Assembloid co-culture systems incorporating cancer-associated fibroblasts, autologous immune cells, and even intraluminal microbial injection are repositioning PDCOs as immuno-avatars capable of modelling the three core immune evasion mechanisms the authors term camouflage, coercion, and cytoprotection.  Both reviews identify standardised QC frameworks, encompassing histopathological concordance, genomic stability, and functional drug-response consistency, as the field’s most urgent unmet need.

AI is moving from analytical tool to essential infrastructure.

Across all five articles, AI features as more than an accessory.  In triple negative breast cancer research (Sani et al.), AI frameworks for image-based phenotyping and morphological characterisation are identified as essential for scaling organoid platforms toward high-throughput screening.  In gastric cancer (Ma et al.), AI-augmented workflows encompassing OrganoID-based organoid tracking, digital twin simulation, and CBAM-YOLOv3-based viability monitoring are positioned as the route to a 10–14-day functional precision oncology pipeline.  In TARGET-SL (Gillman et al.), the computational engine is the screening strategy itself.  And in the clinical trial landscape review (Skowronska et al.), AI-enabled platforms are classified as a defining feature of the third wave of PDO translation.  The convergence of AI with organoid workflows is no longer a future aspiration.  It is actively reshaping how platforms are designed, validated, and deployed.

Standardisation and equity remain structural barriers to impact.

A consistent thread across all five articles is the gap between scientific capability and clinical scalability.  Variable organoid establishment success rates (ranging from 15% in prostate cancer to over 70% in colorectal), reliance on animal-derived Matrigel, turnaround times exceeding six weeks, and the absence of harmonised QC and regulatory frameworks are cited across multiple articles as obstacles to broader adoption.  Ma et al. specifically call for a tripartite validation standard; Skowronska et al. flag geographic concentration of trial activity and equitable access to organoid-guided therapies as concerns.  Sani et al. note that the current clinical cost of personalised organoid testing ($6,000–$12,000+) limits real-world integration.  Together, these articles signal that translational progress will increasingly depend not only on biological innovation, but on platform industrialisation, regulatory harmonisation, and a commitment to making these tools accessible across healthcare systems.

Overall, this week’s selection marks a maturation point.  The question for organoid and MPS technologies is no longer whether they can model human biology with sufficient fidelity.  It is how rapidly the supporting infrastructure, computational, regulatory, and clinical, can be built to match.

Source Articles:

Sani, J. et al. (2026) The Organoid Decade: Leveraging 3D Patient-Derived Organoids to Bridge the Translational Gap in Triple-Negative Breast Cancer: A Systematic Review. Cells 15, 922; https://www.worc.community/documents/the-organoid-decade-leveraging-3d-patient-derived-organoids-to-bridge-the-translational-gap-in-triple-negative-breast-cancer-a-systematic-review

Skowronska, M. et al. (2026) Patient-Derived Organoids in Clinical Medicine: Proven Impact and Future Directions. Organoids 5, 15; https://www.worc.community/documents/patient-derived-organoids-in-clinical-medicine-proven-impact-and-future-directions

Gillman, R. et al. (2026) Advancing Liver Cancer Precision Medicine with TARGET-SL. bioRxiv; https://www.worc.community/documents/advancing-liver-cancer-precision-medicine-with-target-sl

Ma, J. et al. (2026) Patient-derived organoids in gastric cancer: bridging the tumor microenvironment to functional precision oncology. Frontiers in Bioengineering and Biotechnology 14; https://www.worc.community/documents/patient-derived-organoids-in-gastric-cancer-bridging-the-tumour-microenvironment-to-functional-precision-oncology

Yang, P. & Gao, Y. (2026) Decoding the Tumor Microenvironment Chemokine Network: From Immune Evasion to Innovative Multi-Target Therapies. Immunity & Inflammation; https://scienmag.com/decoding-the-tumor-microenvironment-chemokine-network-from-immune-evasion-to-innovative-multi-target-therapies/

Bioprinting Meets Precision Oncology: A High-Throughput Organoid Platform for Head and Neck Cancer

Head and neck squamous cell carcinoma (HNSCC) is one of oncology's most persistent challenges.  Anatomically diverse, molecularly heterogeneous, and stubbornly resistant to the kind of generalised treatment strategies that work for less variable tumour types.  A new study from Lin et al., posted to bioRxiv (DOI: 10.64898/2026.05.20.726741), proposes a practical path toward changing that; not through a new therapy, but through a new way of selecting them.

At the heart of the platform is a bioprinting step that addresses one of the most underappreciated barriers to clinical organoid translation.  Mechanical fragility.  Patient-derived organoids are notoriously delicate, making them difficult to handle at the throughput needed for systematic drug screening.  By printing ring-structured organoids in a defined hydrogel geometry compatible with automated liquid handlers, the authors sidestep this bottleneck entirely, enabling a 33-compound library screen, spanning chemotherapy, targeted agents, and immunotherapy, to be run in parallel across multiple organoid lines with minimal manual intervention.

The screening readout goes further than viability alone.  An ATP-based luminescence assay provides the endpoint measure, but it is supplemented throughout the culture period by a machine learning-driven brightfield image analysis pipeline that tracks organoid size and circularity daily.  This morphological layer proves significant.  It identified that the AKT inhibitor, ipatasertib, promotes ECM invasion in HN30 organoids, a finding that viability alone would have missed and that carries direct implications for treatment selection in a disease context where lymph node metastasis affects a substantial proportion of patients.

Screening across three HNSCC cell line-derived organoids, HPV-negative HN30 and HN31, and HPV-positive SCC154, alongside four HPV-positive patient-derived oropharyngeal tumour organoids, confirmed several subtype-specific patterns consistent with prior literature, including greater radiation sensitivity in HPV-positive lines, while also surfacing clinically actionable radiosensitisers.  Cetuximab, sorafenib, and nedisertib each demonstrated significant radio-sensitising activity across the patient-derived models, with profiles varying meaningfully between individuals; a finding that reinforces the case for functional patient-specific screening over population-level treatment guidelines.

For the microphysiological systems community, the study's most transferable contribution is arguably methodological.  It demonstrates that bioprinting-enabled structural robustness, automated liquid handling, and label-free morphological readouts can be combined into a workflow compatible with clinical timelines and sample constraints.  The platform was validated on primary tumour material collected through existing surgical pathways and processed with standard dissociation protocols, grounding its translational ambitions in applied reality rather than idealised laboratory conditions.

Limitations remain.  The linear interaction model used to identify radiosensitisers loses resolution when a compound is itself highly toxic at screening concentrations, and the circularity metric captures collective invasion well but may not generalise to other modes of cellular dissemination.  Integration with clinical decision timelines, as well as the broader challenge of building outcome-correlated validation datasets, represents the next necessary step.

Taken together, the platform positions patient-derived organoids not as exploratory research tools, but as functional clinical instruments.  Extensions of the patient that enable treatment selection to be tested, refined, and personalised before it is delivered.

Source Article:

Luda Lin, Krishna K. Bommakanti, Christian Wooten, Alfredo Enrique Gonzalez, Yazeed Alhiyari, Jonathan Levi, Bowen Wang, Andreanne Sannajust, Lauran K. Evans, Peyton Tebon, Maie A. St. John & Alice Soragni (2026) A Bioprinted Head and Neck Cancer Organoid-Based Platform for Evaluating Multimodal Therapies. bioRxiv; https://www.worc.community/documents/a-bioprinted-head-and-neck-cancer-organoid-based-platform-for-evaluating-multimodal-therapies

Hope you enjoyed this week's WORC Break!    As always, it would be great to hear your feedback on this feature, share it and join the discussion!  Bookmark this page to stay on top of all things 3D and look forward to connecting with you!

Editor