Improving Antibody–Drug Conjugate Safety Profiling: The Need for Translational Ex Vivo Models to Assess Hematotoxicity
Antibody–drug conjugates (ADCs) hold great promise for targeted cancer therapy, but their clinical success is often limited by hematotoxicity. Traditional models fail to predict these toxic effects with sufficient accuracy, particularly in the human bone marrow. In this post, we explore how Vivia Biotech’s proprietary ex vivo assay provides a translational solution to this challenge. By enabling simultaneous measurement of ADC efficacy and hematopoietic toxicity in a native microenvironment, this platform offers drug developers a powerful tool to optimize the therapeutic index, compare ADC candidates, and anticipate safety liabilities—early and reliably.
Introduction: The Challenge of Dose Limiting Toxicity in ADC Development
Antibody drug conjugates (ADCs) have transformed oncology by enabling targeted delivery of cytotoxic agents to cancer cells. However, despite the clinical success of some ADCs, many fail during development due to toxicity concerns. Hematologic toxicity is one of the most frequent and limiting adverse effects, often determining the maximum dose that can be administered in patients.
This discrepancy between the intended selectivity of ADCs and their observed clinical toxicity reflects a key limitation in current preclinical strategies. Most traditional models fall short in predicting human hematotoxicity with sufficient precision.
Existing Tools to Assess ADC Hematotoxicity: Strengths and Shortcomings
ADC toxicity is typically assessed using a combination of animal models, in vitro cell cultures, and colony forming assays. While each of these approaches contributes valuable information, they all present limitations:
- Animal models often fail to capture the pharmacodynamics and antigen distribution observed in human tissues.
- In vitro cell lines oversimplify the tumor and bone marrow environment.
- Progenitor colony assays do not replicate ADC delivery mechanisms or immune-mediated toxicity.
Importantly, none of these systems allow for simultaneous and physiologically relevant evaluation of efficacy and hematotoxicity in human samples.
Mechanisms Behind ADC Related Hematotoxicity
Recent reviews, such as Nguyen et al. (2023), describe multiple mechanisms contributing to ADC toxicity:
- Off target and off site toxicity due to systemic release of the cytotoxic payload.
- Non specific uptake in immune cells that express Fc receptors, such as megakaryocytes and macrophages.
- On target toxicity when the antigen is also present in healthy hematopoietic cells.
These effects frequently result in neutropenia, thrombocytopenia, or anemia, which compromise the therapeutic window.
The Need for Translational Ex Vivo Models
To overcome these challenges, drug developers increasingly recognize the need for functional human based models that preserve the complexity of the native microenvironment.
An ideal system should:
- Maintain the architecture of the bone marrow niche.
- Quantify selective cytotoxicity on malignant and healthy cells.
- Reproduce immune mediated mechanisms.
- Generate actionable data such as therapeutic index and effective dose ranges.
Measuring ADC-Induced Hematotoxicity Under Realistic and Maximized Conditions
One of the greatest challenges in ADC development is accurately evaluating hematotoxicity, especially when off-target effects occur in the bone marrow. To address this, Vivia Biotech has developed a proprietary ex vivo assay designed to measure simultaneously the antitumor activity and hematopoietic toxicity of ADCs, using patient-derived samples
This assay has been specifically optimized to capture toxicity at its maximum potential. For ADCs targeting solid tumors, cytotoxic activity is typically localized to the tumor site, while hematotoxicity arises from systemic exposure of the payload reaching the bone marrow. Our approach bypasses this spatial separation by co-culturing solid tumor cells directly with healthy human bone marrow (NBM). This allows us to observe, in a single experiment, both the cytotoxic effect on tumor cells and the impact on myeloid progenitors, under identical exposure and microenvironmental conditions.
By removing the variable of payload pharmacokinetics between compartments, the assay provides a conservative and robust estimate of maximal hematotoxicity, making it particularly useful to compare different ADCs side-by-side, or evaluate variations in linker stability and payload release.
In contrast to simplified systems, the assay incorporates the full cellular and soluble complexity of the bone marrow, including autologous plasma, red blood cells, platelets, NK cells and complement proteins. This enables detection of immune-mediated mechanisms such as ADCC or CDC, independent of the cytotoxic payload. These factors are especially relevant for ADCs with Fc-engineered antibodies or in settings where antibody effector function contributes significantly to efficacy or toxicity.
Beyond comparative profiling, this system also enables mechanistic exploration of ADC pharmacology. It can reveal how properties such as internalization rate, linker cleavability or payload diffusibility influence both tumor selectivity and hematopoietic safety. For example, in one comparative analysis, three ADCs sharing the same target and payload but differing in linker stability showed identical tumor activity and progenitor toxicity at 72 hours, suggesting comparable effective release under ex vivo conditions despite linker differences.
In summary, this assay provides a functionally integrated, mechanistically informative, and translationally relevant tool to evaluate ADC-induced hematotoxicity. It complements traditional efficacy screening by adding a critical layer of safety profiling, supporting rational ADC design and de-risking clinical development.
Conclusion
As the field of ADCs continues to expand, a shift toward human relevant preclinical models is essential. Predicting toxicity with accuracy, particularly in hematologic compartments, requires systems that reflect the complexity of patient biology.
Vivia Biotech’s ex vivo platform bridges this gap, offering a high resolution view of therapeutic index under conditions that closely mimic the clinical setting. This approach facilitates smarter compound selection and safer, more effective therapies.
Reference
Nguyen, T., He, J., Sheck, L. H. Q., et al. (2023). Mechanisms of ADC Toxicity and Strategies to Increase ADC Tolerability. Cancers, 15(3), 713.