The in vitro data package: lead selection to IND

What in vitro pharmacology data do you need for a biologic IND filing? A practical guide to mechanism of action, potency, species relevance, immune safety and ICH S6(R1).

An in vitro pharmacology data package is a set of laboratory results, generated in cellular and biochemical systems rather than in animals, that demonstrates what a candidate therapeutic does, how potently, how selectively and with what early signs of risk.

For a therapeutic antibody, bispecific, cell engager, Fc-fusion or other immune-modulating biologic, it starts at lead selection, grows through lead optimisation and candidate selection, and ends up as the pharmacology section, and part of the safety case, of an investigational new drug (IND) application or clinical trial application (CTA).

This guide sets out what belongs in the package at each stage, which assays to run first, what the IND itself needs, how human in vitro assays de-risk a programme and reduce reliance on animal studies, how to show that an animal species is pharmacologically relevant, what investors and pharma partners look for, why packages get questioned, and how to plan the work with an outsourced partner CRO.

What does the pre-clinical pathway look like for a biologic?

The pre-clinical pathway for a therapeutic antibody or immunotherapy runs from target validation through lead selection, lead optimisation and candidate selection to the IND-enabling studies that support a first-in-human trial. In vitro work carries most of the pharmacology at every stage, and the questions it answers change as the programme matures.

The milestones investors and partners recognise sit at the boundaries: a validated target, a selected candidate with a differentiated profile, a complete IND-enabling package and a cleared IND.

Which in vitro assays should you run first?

Run the assays that validate the thesis: does the molecule bind the target on real cells, does that binding do something useful, and does it also bind and act on the target in the species you'll need for toxicology assessments? These are important early questions.

For a new antibody or immunomodulator, a sensible first wave is:

1. Binding affinity and kinetics by surface plasmon resonance (SPR) or bio-layer interferometry (BLI), against the human target and the orthologues from candidate toxicology species.
2. Cell binding by flow cytometry on cells that express the target at physiological levels, not only on over-expressing lines.
3. A functional assay that reports the intended mechanism. For an agonist or antagonist, a reporter cell line for the pathway gives a fast, reproducible first read; for a depleting or cytotoxic molecule, a killing assay with the relevant effector cells.
4. Species cross-reactivity in function, not just binding. Consider which in vivo models will be necessary later down the line. Understanding the requirement for cross-reactivity, including functional activity towards murine or cynomolgus orthologues, up-front is key. A molecule that binds the cynomolgus target but does not produce the expected functional response in cynomolgus cells may limit the value of that species for subsequent pharmacology or toxicology studies and should be understood early in development.

Once leads are shortlisted, lead optimisation ranks them on potency in the relevant cellular system (e.g. primary human immune cells, or appropriate target-expressing cells), selectivity against related targets (biophysical and cell-based analyses), demonstration of Fc effector function that matches the design (silenced, retained or enhanced) and early developability signals such as aggregation and thermal stability.

Ranking on primary cells from several donors at this stage can separate the candidates that survive contact with human biology from those that only work in a cell line.

What in vitro pharmacology does an IND need?

The in vitro data in an IND needs to characterise pharmacology and support species selection and, where relevant, the rationale for first-in-human dose selection. The regulations describe the requirement in general terms, so the content is shaped by guidance, chiefly ICH S6(R1) for biotechnology-derived products, and by what reviewers have come to expect.

In practice, the in vitro elements of an IND pharmacology package for a biologic are:

1. Primary pharmacodynamics. Binding affinity and kinetics, and functional potency in the relevant human cells, with dose-response curves and defined EC50 or IC50 values. This is the evidence that the molecule does what it's claimed to do.
2. Mechanism of action. Assays that show how the effect is produced, for example receptor blockade, cell depletion by ADCC, ADCP or CDC, T cell redirection, or checkpoint blockade.
3. Relevant species justification. Comparative assessment of target binding and functional activity in human and candidate animal-species systems can help establish whether the proposed toxicology species is pharmacologically relevant.
4. Functionality. For immune-modulating biologics, this may involve comparing activity in human and relevant animal-species systems, including primary immune cells or engineered target-expressing cells. Depending on the mechanism of action, relevant endpoints could include target binding, functional potency, cytokine responses, immune-cell activation, depletion, cytotoxicity or Fc-mediated effector function.

5. Tissue cross-reactivity. For monoclonal antibodies and related biologics, binding to a panel of human tissues to look for unexpected off-target binding.

6. Fc-mediated effector function, where the Fc is functional or has been engineered: FcγR and FcRn binding, C1q binding, and cell-based ADCC, ADCP and CDC assays.

7. Immune safety. Cytokine release assays (CRA) in human whole blood or PBMCs, in formats appropriate to the mechanism, and an assessment of immunogenicity risk.

8. Starting dose support. For immunomodulators and agonists, a minimum anticipated biological effect level (MABEL) built from receptor occupancy, in vitro concentration-response in human cells and the comparison of human and animal potency.

What reviewers look for is coherence: the concentrations that work in vitro, the exposures achieved in the toxicology species and the proposed clinical dose should line up into one story.

How do human in vitro assays de-risk a programme?

Human in vitro assays de-risk a programme by testing the biologic on the cells it will meet in patients, before money is spent on in vivo studies in a species where the biology may differ. For immunology programmes in particular, the human immune system is often the better model of itself.

Where they earn their keep:

1. Before in vivo studies. Potency and mechanism in primary human immune cells, across several donors, show whether the effect is real and reproducible. A candidate that fails to demonstrate the expected pharmacology in an appropriate human system may warrant further investigation before progression to in vivo studies.
2. Where animals can't answer the question. Many immune targets (and even cells) have no rodent orthologue, or one the molecule doesn't bind. Human cells, patient-derived cells and ex vivo human tissue explants can answer questions a mouse can't.
3. For safety signals specific to biologics. Human in vitro systems can provide information on immune-mediated risks that may not be adequately characterised in conventional animal models. For immune-activating therapeutics, appropriately designed cytokine release assays using human whole blood or PBMCs can help characterise the potential for cytokine-mediated responses before first-in-human administration.
4. Under the current regulatory direction. Changes to US legislation and FDA policy have expanded the scope for non-animal methods to contribute to nonclinical safety assessment. The FDA Modernization Act 2.0/FDORA framework recognised alternatives to traditional animal studies, including cell-based assays and other New Approach Methodologies (NAMs). In April 2025, the FDA published a roadmap to reduce animal testing in preclinical safety studies, initially focusing on monoclonal antibodies. In September 2026, the FDA further updated its regulations to clarify that non-animal methods can be used where scientifically appropriate. Animal studies remain important where they are scientifically justified, but human-relevant in vitro and ex vivo approaches can increasingly contribute to the evidence supporting clinical development.

For a first-in-class immunomodulator, the ex vivo human assays that strengthen the safety case include whole blood and PBMC cytokine release in more than one format, T cell activation and proliferation across donors with a sufficient range of HLA types, and, where the target is present on healthy tissue, an assessment of activity against primary cells from that tissue.

How do you demonstrate that an animal species is pharmacologically relevant under ICH S6(R1)?

Species relevance for a biotechnology-derived therapeutic should not be based on sequence homology alone. ICH S6(R1) describes sequence comparison as a useful starting point, followed by comparative assessment of target binding and functional activity across species.

Cell-based assays using human and candidate animal-species systems can therefore help determine whether a therapeutic engages its target and produces the expected biological response in the proposed toxicology species.

For immune-modulating biologics, this may involve comparing human and cynomolgus monkey primary immune cells, engineered target-expressing cells or other species-relevant systems for binding, potency, cytokine responses, immune-cell activation, depletion or other mechanism-specific functional endpoints.

What do investors and pharma partners look for?

Investors and partners look for an in vitro package that proves the mechanism works in human cells, shows the molecule is differentiated from what already exists, and reveals no obvious reason it will fail as a drug.

The bar rises with each round: a seed investor wants proof of mechanism, a Series A investor wants a selected candidate with a coherent data story, and a licensing partner wants a package they can take to their own regulatory team without rebuilding it.

The questions that come up in diligence:

• Does it work in human primary cells, not only in engineered lines, and in cells from more than one donor?

• Is the potency comparable with, or better than, the clinical-stage competitor molecule in the same assay? A head-to-head benchmark is worth more than a standalone EC50.

• Is the mechanism selective? Activity against related targets, and in cells that lack the target, should be shown rather than assumed.

• Are there early safety flags? Cytokine release, off-target binding and an immunogenicity assessment answer the question before it's asked.

• Is it developable? Aggregation, stability and manufacturability data belong in the story from lead optimisation onwards.

• Is the data reproducible and documented? Assays with defined acceptance criteria, controls and reports that could go into a submission.

A package that reads as a chain of evidence from mechanism to potency to safety to starting dose is what convinces a partner. Isolated experiments, however good, don't.

Why do in vitro packages get questioned, and what delays an IND?

Packages get questioned when the data doesn't connect to the clinical plan: concentrations that don't relate to expected exposure, a toxicology species chosen without functional evidence that the molecule works in it, or a mechanism claimed but not shown.

Many delays can be traced back to a gap that was visible at candidate selection and left until the submission.

The recurring issues:

• No functional evidence of species relevance. Binding to the cynomolgus target isn't enough; reviewers expect to see that the molecule is also active in cynomolgus cells.

• Concentrations disconnected from exposure. In vitro activity shown only at concentrations far above or below the anticipated clinical range.

• A single donor, or a single cell line. Immune responses vary between people; a result from one donor doesn't establish a human response.

• Assays without controls or acceptance criteria. A potency value with no reference standard, no system suitability criteria and no record of assay performance is hard to rely on.

• Cytokine release tested in one format only. Different formats detect different mechanisms of release, and a negative result in an unsuitable format doesn't clear the risk.

• Immunogenicity left unassessed. Anti-drug antibodies (ADAs) can affect exposure, pharmacology, efficacy and, in some cases, safety. Immunogenicity risk should therefore be considered as part of the development strategy.
• Effector function that doesn't match the design. An Fc claimed as silent needs data showing it's silent. Absence of effector function needs to be demonstrated more robustly than positive effector function, to convince of a true negative. Positive in-assay controls are key.

• Late potency assay development. Leaving development of a relevant biological potency assay until late in the programme can create additional CMC work and make it harder to establish a consistent link between the product's mechanism of action and measures of biological activity.

   

How long does it take, and how do you plan it with a CRO?

With careful planning, a robust in vitro pharmacology data package can, in many cases, be generated within approximately six to twelve months.

The in vitro pharmacology programme supporting an IND typically runs alongside toxicology and CMC activities in the months leading up to submission. Timelines depend on the complexity of the molecule and mechanism of action, the assays required, and how much work completed during lead optimisation can be carried forward.

Assays developed with later-stage requirements in mind can shorten the IND-enabling phase, whereas assays that need to be redesigned or repeated can add time.

Efficiency also comes from matching assay complexity to the stage of development. Early screening may use relatively high-throughput cell-based or reporter assays to compare hundreds of molecules, progressively narrowing these to a smaller number of candidates for more physiologically relevant and complex assays using primary human cells, co-cultures or patient-derived material.

Donor numbers should similarly be matched to the question being asked. Functional primary immune-cell assays can initially be performed across a small number of donors—for example, three or more—to establish activity and donor-to-donor variability, before expanding the cohort where greater biological diversity is important.

In contrast, in vitro T cell assays used for immunogenicity risk assessment typically require substantially larger, appropriately selected donor cohorts to provide representative HLA coverage; cohorts of approximately 30–50 donors are commonly used, depending on the molecule, assay and risk assessment.

Planning the work with a collaborative research partner CRO goes better when the scope is framed as questions to be answered rather than as a list of assays:

1. Start from the claims you'll make in the submission and work back to the data each one needs.

2. Agree the species question early, because it decides the toxicology programme and the timeline.

3. Set donor numbers, comparators and acceptance criteria before the first experiment, so the data is usable in the submission without repetition.

4. Ask for study design as well as execution. A CRO that challenges the plan, suggests the right format for a cytokine release assay or points out a missing control is doing part of the regulatory work for you. It is key that what you are trying to achieve, i.e. the core of the question to be answered, is shared with your CRO partner to allow them to guide you with their experience.
5. Plan the potency assay for lot release from the outset, so it's validated when CMC needs it. Although a validated potency assay may not be required until later clinical stages, understanding early which cell-based system best represents the primary MoA allows the potency assay to be integrated more easily as the candidate biologic is clinically progressed.

RoukenBio works this way as a preclinical immunology CRO. Our in vitro pharmacology, biophysical, immunotoxicology and bioanalytical teams design the study with you, run it in primary human immune cells and patient tissue and reporter systems, and report it in a format that goes straight into the package.

Our preclinical services page sets out the full scope, from discovery and development through to cytokine release and potency assay development.

For antigen-density work on the target cell side, see RoukenCells' guide to antigen density and the therapeutic window.