🎗 Organoid Co-Culture with Immune Cells

The Limitation of Tumor-Only Organoid Culture — No Immune Compartment, No Immunotherapy Signal

Patient-derived tumor organoids (PDTOs) are grown from a biopsy or resected tumor sample, embedded in an extracellular-matrix scaffold, and expanded in a defined growth-factor cocktail. In this standard form, the culture contains only tumor (and sometimes tumor-associated stromal) cells. This makes PDTOs excellent for predicting sensitivity to cytotoxic chemotherapy and targeted agents — the drug acts directly on tumor cells present in the dish. But immunotherapy does not work this way: checkpoint inhibitors, T-cell engagers, and adoptive cell therapies act on the interaction between immune cells and tumor cells. With no immune compartment in the model, there is nothing for these agents to modulate, and the single most transformative class of modern oncology drugs cannot be functionally tested at all.

• Tumor only: Standard PDTO composition (no lymphocyte compartment)
• ~80–90%: Chemo response prediction (concordance in retrospective series)
• Not possible: Immunotherapy prediction (no immune cells to engage)
• 2–6 weeks: Organoid culture window (from biopsy to assay-ready)

Why immune-oncology agents need an immune compartment to be tested

Cytotoxic chemotherapy, and to a large extent targeted small molecules, act through direct, cell-intrinsic mechanisms: a drug diffuses into the culture, engages its molecular target inside or on the tumor cell, and tumor cell death or growth arrest follows. A tumor-only organoid captures this interaction completely — the assay readout (viability, proliferation) directly reflects the drug's mechanism.

Immune checkpoint inhibitors work through an entirely different logic. Drugs such as anti-PD-1 or anti-PD-L1 antibodies do not kill tumor cells directly. They block an inhibitory signal that would otherwise suppress a T-cell that is already trying to recognize and kill the tumor cell. The therapeutic effect only exists in the presence of that T-cell. Remove the T-cell from the system, and an anti-PD-1 antibody added to a tumor-only organoid does precisely nothing — not because the drug fails, but because the assay has no biology left for it to act on.

This is the central rationale for organoid-immune co-culture: to functionally test immunotherapy, the model must contain both halves of the interaction — the tumor cell displaying antigen and immune checkpoint ligands, and the T-cell capable of recognizing and killing it. Building that second compartment into the organoid system is the subject of the remaining stages of this model.

Building the Co-Culture — Adding Autologous or Matched T-Cells to the Organoid System

To create a functional immunotherapy assay, T-cells are introduced directly into the organoid culture well. The most informative configuration uses autologous T-cells — isolated from the same patient's peripheral blood or from tumor-infiltrating lymphocytes (TILs) expanded from the same resected specimen — because these carry the patient's own T-cell receptor repertoire and prior tumor-antigen exposure. Where autologous material is unavailable, HLA-matched or engineered T-cells (e.g., bearing a defined TCR or CAR against a tumor antigen) can substitute, though the biological readout then reflects that specific receptor rather than the patient's endogenous repertoire.

• Autologous PBMC / TIL: Preferred T-cell source (patient-matched repertoire)
• HLA-matched / engineered: Alternative source (defined TCR or CAR)
• 1:1 – 20:1: Typical seeding ratio (T-cell : tumor cell)
• 24–96 h: Co-culture duration (before killing readout)

Sourcing and adding T-cells to an established organoid culture

T-Cell Isolation: • Peripheral blood mononuclear cells (PBMCs) are isolated from a patient blood draw by density-gradient centrifugation. • Tumor-infiltrating lymphocytes (TILs) can instead be expanded directly from a fragment of the resected tumor — these are already tumor-experienced and often enriched for tumor-reactive clones. • T-cells may be used unstimulated, or pre-activated/expanded ex vivo (e.g., with anti-CD3/CD28 beads and IL-2) to increase yield before co-culture.

Adding T-cells to the organoid well: • The organoid, still embedded in its matrix dome or in suspension culture, is combined with a defined number of T-cells pipetted directly into the surrounding medium. • The T-cell : tumor cell ratio (E:T ratio, "effector : target") is the single most important experimental variable — low ratios mimic a sparse immune infiltrate, high ratios probe maximal killing capacity. • Cytokines such as IL-2 or IL-15 are often added to support T-cell survival across the multi-day assay window, since organoid culture medium is not otherwise optimized for lymphocytes.

At this stage the two cell populations simply share a well — the T-cells have not yet penetrated the three-dimensional organoid structure, which is the physical and biological barrier addressed in the next stage.

T-Cell Infiltration — Crossing the Organoid Boundary Is Itself a Measurable Biology

A T-cell floating in the medium around an organoid cannot kill anything until it physically migrates into the three-dimensional tumor mass and makes direct contact with a tumor cell. This infiltration step is not a formality — it recapitulates one of the central determinants of immunotherapy failure in patients: "immune-excluded" tumors, where T-cells surround but never penetrate the tumor, and "immune-desert" tumors, where almost no T-cells are present at all. By tracking how many, and how quickly, T-cells cross the organoid boundary, the co-culture assay provides a direct readout of tumor architecture's permissiveness to immune access — independent of whether killing subsequently occurs.

• ~40–50%: Immune-excluded tumors (of solid tumors, by phenotype)
• Live imaging: Infiltration readout (confocal / lightsheet time-lapse)
• 6–24 h: Time to infiltrate (typical, ratio-dependent)
• ECM density, chemokines: Barrier factors (stroma, physical packing)

What determines whether a T-cell can physically enter the organoid

Chemotactic guidance: • Tumor and stromal cells secrete chemokines (e.g., CXCL9/10/11 acting on CXCR3, or their absence) that either draw T-cells inward or fail to provide a directional cue, leaving T-cells to patrol the periphery indefinitely.

Physical/matrix barriers: • The extracellular-matrix dome the organoid is embedded in, and the density of the tumor cell packing itself, can mechanically impede T-cell migration — dense collagen or a tightly packed epithelial core presents a very different infiltration challenge than a loosely packed, necrotic-cored structure.

Adhesion and integrin engagement: • T-cells use integrins (e.g., LFA-1 binding ICAM-1 on tumor or endothelial-like cells) to grip and crawl through tissue; low ICAM-1 expression on tumor cells correlates with poor T-cell infiltration in both organoids and patient tumors.

Imaging and quantification: • Live confocal or lightsheet time-lapse imaging with fluorescently labeled T-cells (e.g., CellTracker dyes or a constitutive reporter) allows direct visualization and quantification of the fraction of T-cells that cross the organoid boundary versus those that remain at the periphery over the imaging window.

A co-culture that shows abundant T-cells surrounding the organoid but almost none inside it, at any E:T ratio, models an immune-excluded tumor — informative even before any killing assay is run, because it predicts that checkpoint inhibition alone (which only unleashes T-cells already in contact with tumor) is unlikely to help without first addressing the infiltration barrier.

Immune-Mediated Tumor Cell Killing — A Functional Readout of Anti-Tumor Activity

Once T-cells have infiltrated and made contact with tumor cells, the co-culture assay moves to its core functional endpoint: quantifying tumor cell death caused specifically by immune engagement. A T-cell recognizing its cognate antigen on an MHC-presenting tumor cell forms an immunological synapse, releases perforin and granzymes (or engages Fas/FasL), and drives the tumor cell into apoptosis. Measuring how much of this happens — and how fast — gives a direct, patient-specific functional measurement of anti-tumor immune activity that no static biomarker (like PD-L1 immunohistochemistry or tumor mutational burden alone) can fully substitute for.

• Perforin / granzyme: Killing mechanism (+ Fas–FasL apoptosis)
• Live/dead imaging, caspase: Readout methods (organoid viability assays)
• 10–70%: Typical killing range (highly patient-dependent)
• 24–72 h: Assay window (post T-cell engagement)

Quantifying immune-mediated killing in the organoid co-culture

Readout technologies: • Live/dead fluorescent dyes (e.g., a membrane-impermeant DNA dye that only labels dead cells) combined with time-lapse confocal imaging track the fraction of tumor cells dying over the co-culture window. • Caspase-3/7 activation reporters give a real-time apoptosis signal specifically within organoid cells, distinguishable from T-cell death. • Bulk organoid viability assays (e.g., ATP-based luminescence) after a fixed co-culture period give a simpler, plate-based endpoint suited to testing many patient samples or many drug conditions in parallel. • Single-cell dissociation followed by flow cytometry can separately quantify surviving tumor cells, activated T-cells, and exhausted T-cells at assay endpoint.

Why this readout matters clinically: • Two patients with identical PD-L1 expression and tumor mutational burden can show completely different killing in this assay — one organoid is efficiently cleared, the other barely touched — because the functional assay captures the net result of antigen presentation, infiltration, immunosuppressive microenvironment, and T-cell fitness simultaneously, rather than any single biomarker in isolation. • Because the assay uses the patient's own tumor and (ideally) autologous T-cells, the killing percentage is a direct, ex vivo simulation of what checkpoint blockade or adoptive cell therapy might achieve in that specific patient — the premise explored further in the final stage.

Testing Immunotherapy Agents in the Co-Culture — A Functional Prediction of Patient Response

With a working tumor-T-cell co-culture and a validated killing readout in place, the system becomes a platform for testing immunotherapy agents directly. Checkpoint inhibitors (anti-PD-1, anti-PD-L1, anti-CTLA-4), bispecific T-cell engagers, or other immunomodulatory compounds are added to parallel wells, and the assay asks a very specific question: does this agent increase T-cell infiltration into the organoid, and does it increase tumor cell killing, relative to the untreated co-culture? Because the readout comes from the patient's own matched tumor-immune pair, a positive functional result is a direct, ex vivo prediction of that patient's likely response to the same agent given clinically — a capability no static biomarker panel can offer.

• Anti-PD-1/PD-L1, CTLA-4: Agents testable (bispecifics, cytokines)
• Drug vs. vehicle co-culture: Readout comparison (paired, same patient tumor)
• Encouraging, early-stage: Reported concordance (organoid response vs. clinical outcome)
• ~1–3 weeks: Turnaround (biopsy to functional readout)

How checkpoint inhibitor testing is run and interpreted in the co-culture

Experimental design: • The same patient-derived organoid and autologous (or matched) T-cell pool is split across parallel wells: one arm receives vehicle control, the other receives the immunotherapy agent at a clinically relevant concentration, at matched E:T ratio and time point. • Infiltration and killing are measured in both arms using the same imaging and viability methods established in Stages 3 and 4, and the drug effect is expressed as the delta between treated and untreated co-culture.

Mechanistic interpretation: • Checkpoint inhibitors do not directly kill tumor cells or directly cause T-cells to move — they remove a brake. A positive result (higher infiltration and/or higher killing with the drug) indicates the patient's T-cells were present and capable, but partially held back by an inhibitory checkpoint interaction that the drug relieved. • A null result (no change with the drug) can mean several different things depending on which upstream stage failed: if T-cells never infiltrated at all (Stage 3 failure), checkpoint blockade has nothing to unleash; if T-cells infiltrated but showed exhaustion markers rather than a checkpoint-blockable phenotype, a different agent class may be needed.

Why this matters: • This functional, ex vivo readout captures the combined effect of tumor antigenicity, immune infiltration capacity, and checkpoint biology in one measurement from one patient's own tissue — offering a route toward selecting immunotherapy for individual patients based on direct functional evidence rather than population-level biomarker statistics alone.

Because the organoid and the T-cells both come from the same patient, an increase in infiltration and killing after adding a checkpoint inhibitor in this dish is the closest ex vivo analogue available to asking "would this drug work in this specific patient" — turning the co-culture into a prospective, functional companion diagnostic candidate for immunotherapy selection.