📚 Part of the TILA-278 Regulatory Dossier — Reader's Guide. This article shows the live document; edits to the source appear here automatically.

---

Module 2.4 — Nonclinical Overview (TILA-278)

Field	Value
Document ID	M2.4
Version	1.0
Compound	TILA-278 (anti-TL1A antagonist / IL-22R agonist bispecific)
Standard	ICH M4S
Confidentiality	Confidential

CTD summary for the TILA-278 program; clinical figures trace to Study TILA278-201.

Change History

Version	Date	Author	Summary
1.0	2026-07-08	Clinical/Regulatory	Initial issue

2.4.1 Overview of the Nonclinical Testing Strategy

TILA-278 (Virtual Biopharma Inc.) is a humanized bispecific immunoglobulin G1 (IgG1) monoclonal antibody developed for the induction treatment of moderate-to-severe ulcerative colitis (UC). The molecule combines two functionally distinct binding activities in a single antibody: one arm is an antagonist of TL1A (TNFSF15), and the second arm is an agonist of the interleukin-22 receptor (IL-22R). TL1A antagonism is intended to dampen TH1/TH17-driven mucosal inflammation and the fibrotic response that perpetuates intestinal injury, while IL-22R agonism is intended to drive intestinal epithelial regeneration and mucosal-barrier repair. The two activities are mechanistically complementary — one arm removes an inflammatory driver, the other actively restores the epithelium — and the nonclinical program was designed to characterize each activity individually and the combined pharmacology, and to establish an adequate margin of safety to support the doses evaluated in the pivotal Phase 2b induction study, TILA278-201.

The nonclinical strategy was designed and conducted in accordance with ICH S6(R1) (preclinical safety evaluation of biotechnology-derived pharmaceuticals), with study timing relative to the stage of clinical development guided by ICH M3(R2), reproductive assessment per ICH S5(R3), and safety pharmacology per ICH S7A. Because TILA-278 is a monoclonal antibody, several study types that are standard for small molecules are not scientifically warranted and were not conducted: genotoxicity (ICH S2(R1)), standard rodent carcinogenicity bioassays (ICH S1), broad off-target receptor/enzyme profiling, hERG assessment, and a dedicated thorough QT study (ICH E14/S7B and the E14/S7B Q&A waiver rationale for biologics). The justifications for these exclusions are given in the relevant sections below and are consistent with established regulatory practice for therapeutic antibodies.

Species-relevance strategy. The governing scientific challenge for this program is that TILA-278 was engineered against human targets and is human-target-specific. Neither binding arm recognizes the rodent orthologues of its target with functionally meaningful affinity, so conventional rodents (rat, mouse) are not pharmacologically relevant species for either the pharmacology or the toxicology of the clinical molecule. Species selection therefore followed the ICH S6(R1) principle of identifying pharmacologically relevant species in which the antibody engages its targets and elicits a pharmacodynamic response:

• Cynomolgus monkey was confirmed as the single pharmacologically relevant non-rodent species. Both arms of TILA-278 cross-react with the cynomolgus orthologues of TL1A and IL-22RA1 with affinity and functional potency within approximately 2- to 3-fold of the human targets, and functional agonism at cynomolgus IL-22R and antagonism of cynomolgus TL1A signalling were confirmed in vitro. The cynomolgus monkey was therefore used for repeat-dose toxicology, toxicokinetics, integrated safety pharmacology, and reproductive assessment.
• Rodent disease pharmacology could not be studied with the clinical molecule. In vivo proof of mechanism was therefore generated with a homologous murine surrogate bispecific antibody (mouse-cross-reactive anti-TL1A antagonist / mouse IL-22R agonist) evaluated in established mouse models of colitis. Where a rodent-based hazard assessment was scientifically necessary (e.g., aspects of reproductive toxicity), the surrogate was likewise used.
• Tissue cross-reactivity (TCR) studies using full human and cynomolgus tissue panels were performed for both binding specificities to confirm that the pattern of binding is restricted to expected target-expressing tissues and to bridge the toxicology species to humans.

This dual approach — the clinical molecule in the cross-reactive cynomolgus monkey for toxicology and toxicokinetics, and a homologous surrogate in rodent disease models for in vivo pharmacology — is the standard ICH S6(R1) solution for a human-specific biologic and provides the interpretive basis for the integrated assessment in Section 2.4.5.

Table 2.4.1-1. Overview of the Nonclinical Program

Domain	Study type	Test system	Route	GLP	CTD (M4)
Primary PD (in vitro)	Binding, TL1A neutralisation, IL-22R agonism, selectivity, dual engagement	Human & cynomolgus recombinant proteins/cells; primary human intestinal epithelial cells	—	Non-GLP	4.2.1.1
Primary PD (in vivo)	Colitis models with murine surrogate	Mouse (adoptive T-cell transfer, DSS)	IP/SC	Non-GLP	4.2.1.1
Secondary PD	Tissue cross-reactivity (both arms)	Human & cynomolgus tissue panels	—	GLP	4.2.1.2 / 4.2.3.7
Safety pharmacology	Cardiovascular / respiratory / CNS (integrated); in vitro cytokine-release assay	Cynomolgus; human whole blood/PBMC	SC / —	GLP / Non-GLP	4.2.1.3
PK / TK	SC pharmacokinetics, TMDD, ADA impact	Cynomolgus	SC	Non-GLP / GLP	4.2.2
Repeat-dose toxicity	4-week and 26-week, with recovery	Cynomolgus	SC	GLP	4.2.3.2
Reproductive/developmental	ePPND (+ surrogate fertility/EFD as needed)	Cynomolgus (+ mouse surrogate)	SC	GLP	4.2.3.5
Local tolerance	Injection-site evaluation (integrated)	Cynomolgus	SC	GLP	4.2.3.6
Genotoxicity	Not conducted (scientifically unwarranted)	—	—	—	4.2.3.3
Carcinogenicity	Not conducted (weight-of-evidence assessment)	—	—	—	4.2.3.4

2.4.2 Pharmacology

2.4.2.1 Primary Pharmacodynamics

In vitro characterization. TILA-278 was characterized for each binding activity and for simultaneous dual engagement. The anti-TL1A arm binds human TL1A with high affinity (K_D ≈ 0.2 nM) and neutralizes TL1A–DR3 (death receptor 3, TNFRSF25) signalling with an IC₅₀ of approximately 0.3 nM in a cell-based reporter assay, blocking the TL1A-driven co-stimulation that amplifies TH1 and TH17 effector responses. The IL-22R agonist arm binds the human IL-22RA1 subunit (the ligand-binding chain of the heterodimeric IL-22 receptor, which signals in complex with IL-10RB) and acts as a functional agonist, inducing STAT3 phosphorylation with an EC₅₀ of approximately 0.5 nM in an epithelial reporter system and driving expression of antimicrobial peptides (e.g., REG3 family, S100A8/A9), mucins, and tight-junction components in primary human intestinal epithelial cells. A bridging (dual-binding) assay confirmed that a single TILA-278 molecule can engage TL1A and IL-22RA1 simultaneously. These activities correspond to the two release/characterization potency assays defined for the drug substance (anti-TL1A neutralisation and IL-22R agonism), providing a direct link between the potency-controlled clinical material and the pharmacology described here.

Selectivity. TILA-278 showed no functionally relevant binding to related TNF-superfamily ligands (including TNF-α, FasL, and LIGHT) or to IL-10-family receptor components other than IL-22RA1, and the tissue cross-reactivity studies (Section 2.4.2.2) detected binding only in expected target-expressing tissues. The IL-22R agonist activity is inherently restricted to IL-22RA1-expressing non-haematopoietic tissues (predominantly epithelial: intestine, skin, liver, pancreas, kidney, lung), a distribution that informed the toxicological attention to epithelial proliferation described in Section 2.4.4.

In vivo proof of mechanism. Because the clinical molecule does not engage rodent targets, in vivo efficacy was established with the homologous murine surrogate bispecific in mouse colitis models. In the adoptive CD4⁺ T-cell transfer model and the dextran sulfate sodium (DSS) model, the surrogate produced dose-dependent reductions in colonic histopathology score (on the order of 50–60% versus vehicle), reduced colonic IL-17 and IFN-γ, increased epithelial regeneration markers (REG3γ, mucins), and improved barrier integrity as measured by reduced FITC-dextran permeability. Head-to-head comparison against matched single-activity surrogate antibodies (anti-TL1A antagonism alone and IL-22R agonism alone) demonstrated that the dual-active molecule produced greater mucosal healing than either activity alone at equivalent exposure, supporting the complementary dual-mechanism hypothesis. These data are the mechanistic and translational basis for the dose-ordered clinical efficacy subsequently observed in TILA278-201 (LS-mean change in modified Mayo at Week 12 of −3.36 [High] and −2.76 [Low] versus −1.00 [Placebo]).

2.4.2.2 Secondary Pharmacodynamics

GLP tissue cross-reactivity studies were conducted for both binding specificities on full panels of human and cynomolgus tissues. Staining was consistent with the known distribution of TL1A (endothelium, activated antigen-presenting cells) and IL-22RA1 (epithelial tissues) and revealed no unexpected off-target binding. The concordance between the human and cynomolgus staining patterns further supports the relevance of the cynomolgus monkey as the toxicology species.

2.4.2.3 Safety Pharmacology

Consistent with ICH S6(R1), safety pharmacology endpoints were integrated into the repeat-dose cynomolgus toxicity program rather than conducted as stand-alone small-molecule studies. Cardiovascular parameters (arterial blood pressure, heart rate, and ECG including QTc, by jacketed/implanted telemetry), respiratory function, and CNS/neurobehavioural observations were evaluated within the 4-week study and showed no adverse effects up to the highest dose tested. As a large protein that does not interact with cardiac ion channels, TILA-278 has no expected hERG liability, and hERG testing was not performed. A dedicated thorough QT study was not conducted; the absence of a plausible direct electrophysiological mechanism for an antibody, together with the negative integrated cardiovascular telemetry, supports the ICH E14/S7B (and E14/S7B Q&A) waiver rationale for this biologic.

Because one arm is an agonist and the molecule is immunomodulatory, an in vitro cytokine-release assay (CRA) using human whole blood and peripheral blood mononuclear cells was performed to assess the potential for cytokine-release syndrome. No meaningful induction of pro-inflammatory cytokines was observed under either soluble or plate-immobilized conditions, indicating a low cytokine-release risk and supporting the first-in-human starting-dose rationale (Section 2.4.5).

2.4.2.4 Pharmacodynamic Drug Interactions

No dedicated pharmacodynamic interaction studies were conducted. As an immunomodulatory antibody, TILA-278 has the theoretical potential to alter cytochrome P450 activity indirectly through the resolution of inflammatory cytokine burden (a class consideration for therapeutic proteins that modulate cytokines); this is addressed as a therapeutic-protein drug-interaction consideration in the clinical pharmacology assessment rather than in a nonclinical study.

2.4.3 Pharmacokinetics

The nonclinical pharmacokinetics of TILA-278 are those expected for a humanized IgG1 antibody and were characterized in the cynomolgus monkey, the pharmacologically relevant species. Key features are summarized in Table 2.4.3-1.

Absorption. Following subcutaneous (SC) administration — the intended clinical route — TILA-278 was absorbed with an estimated bioavailability of approximately 65–75% and a time to maximum concentration of several days, consistent with lymphatic-mediated absorption of a large protein.

Distribution. The steady-state volume of distribution was low (≈ 60–80 mL/kg), consistent with the limited extravascular distribution typical of IgG antibodies.

Metabolism and excretion. TILA-278 is eliminated by catabolism to peptides and amino acids and via target-mediated drug disposition (TMDD); it is not a substrate for cytochrome P450 enzymes and is not eliminated intact by the kidney. Systemic exposure was nonlinear (more-than-dose-proportional) at low doses, where target-mediated clearance predominates, and became approximately dose-proportional at the higher pharmacological and toxicological doses, where the TMDD pathway is saturated. The terminal half-life at pharmacological exposures was approximately 10–14 days, and weekly SC dosing produced modest accumulation (≈ 2–3-fold).

Bispecific considerations. Because each arm can engage a distinct target with its own expression and turnover, the disposition of the two activities was assessed together; both binding activities were retained over the systemic residence of the molecule, and no evidence of preferential clearance of one arm (e.g., through target-mediated internalization) that would produce a functionally monospecific circulating species was observed.

Immunogenicity in the toxicology species. Anti-drug antibodies (ADA) were detected in a proportion of dosed animals (≈ 20–40% incidence), as is expected when a humanized antibody is administered to a non-human species; ADA reduced exposure in a subset of animals but did not preclude maintenance of adequate, interpretable exposure at the top doses across the dosing period. ADA findings in animals are not predictive of human immunogenicity and are used only to confirm the validity of the toxicokinetic exposures underpinning the safety margins.

Placental transfer. As an IgG1, TILA-278 is expected to cross the placenta via FcRn-mediated transport, with minimal transfer in the first trimester increasing through the second and third trimesters; measurable drug in infant serum in the enhanced pre- and postnatal development study (Section 2.4.4.5) confirmed this transfer and is relevant to the reproductive risk assessment.

Table 2.4.3-1. Principal Pharmacokinetic Parameters (Cynomolgus Monkey, SC)

Parameter	Value
SC bioavailability	≈ 65–75%
Terminal half-life (t½)	≈ 10–14 days
Steady-state volume of distribution (V_ss)	≈ 60–80 mL/kg
Clearance	Low; nonlinear at low doses (TMDD), linear at high doses
Accumulation (weekly SC)	≈ 2–3-fold
Metabolic route	Catabolism to peptides/amino acids; non-CYP
ADA incidence	≈ 20–40% (reduced exposure in a subset; top-dose exposure maintained)

2.4.4 Toxicology

2.4.4.1 Single-Dose Toxicity

No stand-alone single-dose toxicity study was conducted. Acute tolerability was assessed within the initial dosing phases of the repeat-dose program, consistent with ICH M3(R2) and S6(R1).

2.4.4.2 Repeat-Dose Toxicity

Repeat-dose toxicity was evaluated in the cynomolgus monkey by the clinical (SC) route in a 4-week study with an 8-week recovery phase and a pivotal 26-week study with a 12-week recovery phase, each at doses of approximately 10, 30, and 100 mg/kg administered once weekly. TILA-278 was well tolerated in both studies. Findings were limited to minimal, reversible injection-site reactions and to changes attributable to the intended exaggerated pharmacology — principally reversible shifts in circulating T-cell subsets consistent with TL1A pathway antagonism — without associated functional immunodeficiency at the exposures tested.

Given the IL-22R agonist activity, particular histopathological attention was directed to IL-22RA1-expressing epithelial tissues (gastrointestinal tract, skin, liver, pancreas, kidney, lung) to detect any adverse epithelial hyperplasia or dysplasia; none was observed at any dose. Given the immunomodulatory TL1A antagonism, animals were monitored for opportunistic and reactivated infections; none attributable to treatment were identified. There was no adverse target-organ toxicity, no adverse effect on clinical pathology beyond the expected pharmacology, and all findings resolved during the recovery periods. The no-observed-adverse-effect level (NOAEL) was the highest dose tested (≈ 100 mg/kg/week) in both studies. The exposures at the NOAEL and the resulting clinical safety margins are presented in Section 2.4.5.

Table 2.4.4-1. Pivotal Repeat-Dose Toxicity Studies (Cynomolgus Monkey, SC)

Study	Duration / recovery	Doses (mg/kg/week)	NOAEL	Principal findings
4-week (with integrated safety pharmacology)	4 wk / 8 wk recovery	10 / 30 / 100	100 (high dose)	Minimal reversible injection-site reactions; reversible pharmacodynamic changes in T-cell subsets; no target-organ toxicity
26-week (pivotal)	26 wk / 12 wk recovery	10 / 30 / 100	100 (high dose)	As above; no adverse epithelial hyperplasia; no treatment-related infection; full reversibility

2.4.4.3 Genotoxicity

Genotoxicity studies were not conducted. As a large recombinant protein composed of natural amino acids, TILA-278 cannot interact directly with DNA, and genotoxicity testing of monoclonal antibodies is not scientifically warranted per ICH S6(R1) and S2(R1).

2.4.4.4 Carcinogenicity

Standard rodent carcinogenicity bioassays were not conducted, consistent with ICH S1 and S6(R1) for a monoclonal antibody and precluded in any case by the absence of a pharmacologically relevant rodent species. The carcinogenic potential was addressed by a weight-of-evidence assessment. Two mechanism-based theoretical considerations were evaluated: (i) sustained IL-22R agonism could, in principle, promote epithelial proliferation; and (ii) TL1A antagonism is immunomodulatory, and prolonged immunosuppression can be associated with malignancy risk as a class consideration. Against these, the 26-week study showed no epithelial hyperplasia, dysplasia, or pre-neoplastic change in IL-22RA1-expressing tissues at any dose, the immunomodulation observed was partial and reversible without functional immunodeficiency, and the current clinical use is a time-limited (12-week) induction regimen. On balance, the nonclinical data do not indicate a carcinogenic hazard; malignancy and serious/opportunistic infection are nonetheless carried as monitored risks in the clinical program and risk-management plan.

2.4.4.5 Reproductive and Developmental Toxicity

Reproductive and developmental toxicity was assessed in an enhanced pre- and postnatal development (ePPND) study in the cynomolgus monkey, the design recommended by ICH S5(R3) and S6(R1) for an IgG antibody that crosses the placenta. Dams were dosed SC from early gestation through parturition. There were no adverse effects on maternal health, no increase in abortion or embryo-fetal loss, no external/visceral/skeletal malformations, and normal infant growth, morphology, and immune-system development through the postnatal follow-up period at exposures up to the high dose. Drug was measurable in infant serum, confirming FcRn-mediated placental transfer. Aspects of fertility and early embryonic development that require a rodent test system were addressed using the murine surrogate, with no adverse effects on mating, fertility, or early embryonic development. The developmental exposure margin is summarized in Section 2.4.5.

2.4.4.6 Local Tolerance

Local tolerance at the SC injection site was evaluated within the repeat-dose studies using the clinical formulation (buffered solution containing polysorbate). Injection-site reactions were minimal and reversible, consistent with the SC route and the antibody class, and are concordant with injection-site reactions being the principal drug-attributable finding in TILA278-201.

2.4.4.7 Immunotoxicity

Immunotoxicity endpoints were integrated into the repeat-dose program and included immunophenotyping of circulating lymphocyte subsets, serum immunoglobulins, cytokine profiling, and a T-cell-dependent antibody response (TDAR) assessment. Findings were limited to the expected, reversible pharmacodynamic modulation of T-cell responses arising from TL1A antagonism, without evidence of unintended immunosuppression, immunostimulation, hypersensitivity, or autoimmunity beyond the intended mechanism. These data support routine clinical monitoring for infection rather than indicating an unexpected immunotoxic hazard.

2.4.4.8 Impurities and Excipients

The drug product is a sterile SC solution formulated in a physiological buffer with polysorbate; all excipients are of established SC safety and used within qualified ranges. Product-related variants relevant to a bispecific IgG1 (aggregates, charge variants, glycosylation species) are controlled by the drug-substance specifications, and the batches used in the pivotal toxicology studies were representative of, and bracketed, the impurity/variant profile of the clinical material, thereby qualifying those attributes at the exposures achieved.

2.4.5 Integrated Overview and Conclusions

The nonclinical package for TILA-278 is internally coherent and adequate under ICH S6(R1) and M3(R2) to support the doses and duration of the Phase 2b induction study TILA278-201, and it directly rationalizes the clinical findings.

Pharmacology and mechanism. In vitro data confirm a clean, potent, and selective dual profile — high-affinity TL1A antagonism and functional IL-22R agonism co-resident in one molecule — and in vivo studies with a homologous surrogate demonstrate that the combined activity reduces colitis severity and restores the mucosal barrier more effectively than either activity alone. This mechanistic package is the nonclinical foundation for the dose-ordered efficacy observed clinically (Week-12 clinical remission 37.3% [High] and 16.2% [Low] versus 0.7% [Placebo]) and for the accompanying dose-ordered reductions in the objective inflammatory biomarkers (C-reactive protein and faecal calprotectin) and the endoscopic improvement recorded in TILA278-201.

Species strategy. The human-target specificity of TILA-278 was addressed by a defensible two-part relevance strategy: the cross-reactive cynomolgus monkey served as the single relevant species for toxicology and toxicokinetics, and a homologous murine surrogate enabled in vivo disease pharmacology and rodent-dependent reproductive endpoints that the clinical molecule could not itself address. Tissue cross-reactivity bridges the cynomolgus data to humans. This is the standard and appropriate resolution for a human-specific biologic.

Safety and exposure margins. TILA-278 was well tolerated in cynomolgus monkeys for up to 26 weeks by the clinical route, with the NOAEL at the highest dose tested and findings confined to reversible exaggerated pharmacology and minor injection-site reactions; the specific theoretical concerns raised by the mechanism — epithelial proliferation from IL-22R agonism and immunosuppression from TL1A antagonism — were specifically examined and not observed at adverse levels. On this basis the first-in-human starting dose was selected using a minimum anticipated biological effect level (MABEL) approach appropriate for an immunomodulatory molecule bearing an agonist arm, and the doses subsequently evaluated in TILA278-201 remain within the exposure boundary established by the pivotal toxicology study, as summarized below.

Table 2.4.5-1. Exposure Margins at the Nonclinical NOAEL Relative to the Clinical High Dose (TILA278-201)

Pivotal study	Basis	NOAEL	AUC margin vs clinical High dose	C_max margin
26-week cynomolgus (SC)	General toxicology	≈ 100 mg/kg/week (high dose)	≈ 20×	≈ 15×
4-week cynomolgus (SC)	General toxicology / safety pharmacology	≈ 100 mg/kg/week (high dose)	≈ 20×	≈ 15×
ePPND cynomolgus (SC)	Developmental	High dose	≈ 10–15×	≈ 10×

The approximately 20-fold AUC margin at the general-toxicology NOAEL comfortably encompasses both the High and Low doses studied in TILA278-201 and supports the clinical exposures achieved, while the developmental margin supports the pregnancy risk characterization in labelling and the risk-management plan.

Translational risk assessment. The nonclinical findings translate into a limited and manageable set of clinical risks that are monitored in the clinical program: injection-site reactions (the principal drug-attributable finding both nonclinically and in TILA278-201), infection risk inherent to an immunomodulatory mechanism, and immunogenicity (assessed clinically by a validated tiered ADA assay with evaluation of impact on PK, efficacy, and safety). Genotoxic, carcinogenic, and cardiac-repolarization hazards are not indicated for this molecule, consistent with its class. In aggregate, the pharmacology establishes the dual anti-inflammatory/mucosal-healing mechanism, the species strategy provides a valid translational bridge, the pharmacokinetics are well understood and antibody-typical, and the toxicology defines an adequate safety margin. The nonclinical package supports the benefit-risk basis for TILA-278 induction dosing in moderate-to-severe ulcerative colitis and for continued clinical development.

2.4.6 List of Literature References

1. ICH Harmonised Tripartite Guideline S6(R1). Preclinical Safety Evaluation of Biotechnology-Derived Pharmaceuticals. 2011.
2. ICH Harmonised Tripartite Guideline M3(R2). Guidance on Nonclinical Safety Studies for the Conduct of Human Clinical Trials and Marketing Authorization for Pharmaceuticals. 2009.
3. ICH Harmonised Guideline S5(R3). Detection of Reproductive and Developmental Toxicity for Human Pharmaceuticals. 2020.
4. ICH Harmonised Tripartite Guideline S7A. Safety Pharmacology Studies for Human Pharmaceuticals. 2000.
5. ICH Guidelines S7B and E14, and the E14/S7B Implementation Working Group Questions & Answers. Clinical and Nonclinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential. 2005; Q&A 2022.
6. ICH Harmonised Tripartite Guidelines S1A/S1B (Carcinogenicity) and S2(R1) (Genotoxicity).
7. European Medicines Agency. Guideline on Strategies to Identify and Mitigate Risks for First-in-Human and Early Clinical Trials with Investigational Medicinal Products (EMA/CHMP/SWP/28367/07 Rev. 1). 2017.
8. Literature on the TL1A (TNFSF15)–DR3 axis in inflammatory bowel disease and intestinal fibrosis, and on IL-22-mediated epithelial regeneration and mucosal-barrier repair, supporting the dual-mechanism rationale for TILA-278.

Guideline references (items 1–7) are the current internationally harmonized standards under which this program was designed and conducted. The individual study reports underlying the nonclinical values summarized in this overview are provided in Module 4 (4.2.1–4.2.3), with tabulated study summaries in Module 2.6.