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

---

SEND / Nonclinical Data Reviewer's Guide (TILA-278)

Document ID: SEND-001
Version: 1.0
Change History: 1.0 — Initial issue.
Standard(s): CDISC SEND

SEND / Nonclinical Data Reviewer's Guide — TILA-278

Orients the reviewer to the standardised nonclinical (SEND) datasets for TILA-278: the studies represented, dataset structure and controlled terminology, and conformance notes, providing traceability from the toxicology study reports to the submitted data. CDISC SEND; ICH.

TILA-278 is a humanized IgG1 bispecific monoclonal antibody produced in a Chinese hamster ovary (CHO) cell culture process and administered by subcutaneous (SC) injection. One arm of the molecule antagonizes TL1A (TNF-like ligand 1A; anti-inflammatory and anti-fibrotic effect) while the second arm agonizes the IL-22 receptor (IL-22R; epithelial and mucosal healing effect), giving a single agent that both dampens mucosal inflammation and promotes barrier restitution in moderate-to-severe ulcerative colitis. Because these dual pharmacologic activities and the SEND datasets that support them are governed by the biologics framework, this guide should be read together with the ICH S6(R1) nonclinical rationale, the ICH Q5A(R2), Q5C, and Q6B quality documentation, and the Module 4.2 study reports to which the datasets provide traceability. The application is submitted as a Biologics License Application under 21 CFR Part 601.

1. Introduction and Scope

This Nonclinical Study Data Reviewer's Guide (nSDRG) describes the CDISC SEND datasets provided in eCTD Module 4 for TILA-278. Its purpose is to give the nonclinical reviewer a single orientation document that (a) enumerates the toxicology and toxicokinetic studies represented in SEND, (b) explains how the standardized datasets were derived from the source study reports, (c) documents the standards and controlled terminology versions used, and (d) summarizes data conformance and any known issues so that findings in the tabulated data can be reconciled against the corresponding study report tables, figures, and listings.

The guide follows the PhUSE nSDRG template structure. It is descriptive rather than interpretive: safety conclusions, no-observed-adverse-effect-level (NOAEL) determinations, and exposure margins are established in the individual study reports and in the Nonclinical Overview (Module 2.4) and Nonclinical Written and Tabulated Summaries (Module 2.6). Where this guide restates a design attribute (species, route, dose levels, duration), it does so only to help the reviewer navigate the datasets.

A note on program design that shapes the dataset inventory: TILA-278 is a monoclonal antibody, and its nonclinical package follows ICH S6(R1). Consistent with that guidance and with established regulatory practice for this modality, the program does not include standard battery genotoxicity studies, a rodent or transgenic carcinogenicity bioassay, an in vitro hERG (IKr) assay, or a dedicated thorough-QT-type cardiovascular study. These assessments are not warranted for a large-molecule antibody that does not distribute intracellularly, is not expected to interact with DNA, and is cleared by proteolytic catabolism and target-mediated pathways rather than by small-molecule metabolism. Cardiovascular, respiratory, and central nervous system safety pharmacology endpoints were instead integrated into the repeat-dose general toxicology studies in the pharmacologically relevant species, in accordance with ICH S6(R1) and ICH S7A. Reviewers should therefore expect no SEND datasets for genetic toxicology, carcinogenicity, or stand-alone in vitro cardiac assays; their absence is by design and is documented in the Nonclinical Overview.

2. Nonclinical Program and Studies Represented in SEND

Selection of the test species was driven by pharmacologic relevance rather than by convention. Binding and functional cross-reactivity of both the anti-TL1A antagonist arm and the IL-22R agonist arm were confirmed in the cynomolgus monkey (Macaca fascicularis) but not in rodents, and the relevant epitopes are conserved between cynomolgus monkey and human. In accordance with ICH S6(R1), the cynomolgus monkey is therefore the sole pharmacologically relevant toxicology species, and the pivotal general-toxicology program was conducted exclusively in that species by the intended clinical (subcutaneous) route.

The following GLP studies are represented in the submitted SEND datasets. Study identifiers (STUDYID) as they appear in the Trial Summary (TS) and Demographics (DM) domains are given in the first column.

STUDYID	Study type	Species / route	Design	Dose groups (mg/kg, SC once weekly)	Recovery
TILA278-NC-101	4-week repeat-dose general toxicity with integrated safety pharmacology and toxicokinetics	Cynomolgus monkey / SC	4/sex/group main; 2/sex/group control and high-dose recovery	0 (vehicle), 10, 30, 100	4 weeks
TILA278-NC-102	13-week repeat-dose general toxicity with integrated safety pharmacology and toxicokinetics	Cynomolgus monkey / SC	5/sex/group main; 2/sex/group control and high-dose recovery	0 (vehicle), 10, 30, 100	8 weeks
TILA278-NC-103	26-week chronic repeat-dose general toxicity with integrated safety pharmacology and toxicokinetics	Cynomolgus monkey / SC	5/sex/group main; 3/sex/group control and high-dose recovery	0 (vehicle), 10, 30, 100	12 weeks
TILA278-NC-201	Enhanced pre- and postnatal development (ePPND) study	Cynomolgus monkey / SC	Group-housed pregnant animals dosed from confirmation of pregnancy through delivery	0 (vehicle), 30, 100	Infant follow-up to postnatal day 180

The vehicle (0 mg/kg) control received the formulation buffer used for the drug product. Dose levels of 10, 30, and 100 mg/kg once weekly bracket the intended clinical exposure with margins established on a toxicokinetic (AUC/Cmax) basis in Module 2.6; 100 mg/kg once weekly was the high dose in the general-toxicology studies and was identified as the NOAEL in the pivotal studies. The chronic 26-week study (TILA278-NC-103) supports the duration of chronic dosing anticipated for maintenance therapy in ulcerative colitis.

Two additional nonclinical assessments are part of the Module 4 package but are not provided as SEND tabulations because their endpoints are not amenable to the SEND findings model: the GLP tissue cross-reactivity (TCR) studies, in which TILA-278 was applied to full panels of human and cynomolgus monkey tissues and evaluated by immunohistochemistry, and the in vitro pharmacology/functional characterization work. These are reported in full in Module 4.2 and are cross-referenced here so that the reviewer does not expect a corresponding dataset.

3. Study Data Standards and Dictionary Inventory

All datasets conform to the CDISC Standard for the Exchange of Nonclinical Data (SEND). The versions used across the represented studies are inventoried below; a single consistent version set was applied to all studies in this submission to simplify cross-study review.

Component	Version / package
SEND Implementation Guide	SENDIG v3.1
Dataset metadata (data definition)	Define-XML v2.0, with the CDISC define2-0-0 stylesheet
Transport format	SAS Transport Format (XPORT / XPT) v5
Controlled Terminology	CDISC SEND Controlled Terminology, quarterly package as identified in each define.xml
Conformance engine	Pinnacle 21 Community, SEND rule set current at time of conversion

Because SEND is a nonclinical standard, the datasets do not use MedDRA or WHO Drug Dictionary. Anatomic, pathology, and finding terms are governed by CDISC SEND Controlled Terminology; where a controlled term did not exist for a given collected result, the verbatim/original text is preserved in the ORRES/original-result variables and a sponsor-defined term is documented in the define.xml value-level metadata. The exact CDISC SEND CT package date applied to each study is recorded in the TS domain (parameter TSPARMCD = "SNDCTVER" style metadata) and in the define.xml, and is identical across the four studies in this submission.

4. Source Data, Data Flow, and Traceability

The flow from source to submission is linear and fully traceable:

1. Source. In-life, clinical pathology, and anatomic pathology data were collected at the testing facility in a validated data-acquisition and pathology system (Provantis-class LIMS). Toxicokinetic bioanalysis (drug concentration) and anti-drug antibody (ADA) results were generated in validated ligand-binding assays under GLP.
2. Study report. Data were summarized, audited, and reported in the GLP final study reports located in Module 4.2. These reports are the authoritative source; the SEND datasets are a standardized representation of the same underlying data.
3. SEND conversion. The contract laboratory's data-standards group mapped the source data to SENDIG v3.1 domains, applied CDISC SEND Controlled Terminology, and generated define.xml v2.0.
4. Validation. Converted datasets were validated with Pinnacle 21; the conformance report and the disposition of each message are summarized in Section 9 and provided with the submission.
5. Submission. XPT datasets, define.xml, the stylesheet, and this reviewer's guide were assembled under the Module 4 datasets folder.

Traceability between a tabulated result and its study-report counterpart is preserved through the standard SEND keys: STUDYID identifies the study, USUBJID identifies the animal, and the domain-specific sequence and grouping variables (e.g., --SEQ, SETCD, ARMCD, VISITDY, --DY study day) allow any record to be located against the report's summary tables and individual-animal listings. No derivations that alter reported values were introduced during conversion; derived timing and study-day variables follow the SENDIG algorithms relative to the first dose (reference start date, RFSTDTC).

5. SEND Domain Inventory and Dataset Structure

The domains submitted across the four studies are listed below. Not every domain is populated in every study; the define.xml identifies, per study, which datasets are present and their record counts. Findings domains carry the standard status/exclusion, timing, and result-qualifier variables.

Trial Design (study-level) domains

• TS — Trial Summary: one record per study-level parameter (species, route, dose, duration, standards versions, sponsor, test article identity).
• TA / TE / TX — Trial Arms, Trial Elements, Trial Sets: encode the control and 10/30/100 mg/kg arms, the acclimation/treatment/recovery elements, and the set structure (main study vs. recovery, sex).

Special-purpose domains

• DM — Demographics: one record per animal (age at start, sex, species/strain, arm assignment).
• SE — Subject Elements: the sequence of trial elements actually experienced by each animal.
• CO — Comments: free-text comments linked to animals or records.

Interventions

• EX — Exposure: the SC doses actually administered (dose amount, units in mg/kg, frequency once weekly, route, dose date/day). Injection-site information supports local-tolerance evaluation.

Events

• DS — Disposition: scheduled terminal and recovery sacrifices and any unscheduled removals, with reason.

Findings domains

• BW — Body Weights; BG — Body Weight Gains.
• CL — Clinical Observations (in-life signs, including injection-site observations).
• FW — Food and Water Consumption (as collected).
• LB — Laboratory Test Results (hematology, coagulation, clinical chemistry, urinalysis), including immunophenotyping/lymphocyte-subset and cytokine-related clinical-pathology results relevant to an immunomodulatory antibody.
• MA — Macroscopic (gross) Findings; MI — Microscopic (histopathology) Findings; OM — Organ Measurements (organ weights and ratios); PM — Palpable Masses.
• VS — Vital Signs; and for the integrated safety-pharmacology endpoints, the cardiovascular (CV/EG), respiratory (RE), and nervous-system/neurobehavioral (NV) findings as collected within the repeat-dose studies.
• PC — Pharmacokinetic Concentrations (TILA-278 serum concentrations); PP — Pharmacokinetic Parameters (derived TK parameters).

Relationship and supplemental datasets

• RELREC — inter-record relationships (e.g., linking a PC concentration series to its EX dosing event, or an MI finding to the corresponding MA finding).
• SUPP-- / SUPPQUAL — supplemental qualifiers carrying non-standard variables (for example ADA-related and injection-site qualifiers).
• POOLDEF — pool definitions where any results were collected on pooled specimens.

Anti-drug antibody (immunogenicity) results warrant a specific navigation note. SENDIG v3.1 does not define a dedicated immunogenicity findings domain; ADA incidence, titer, and the neutralizing-antibody determinations for TILA-278 are reported in full in the study reports and, where represented in the datasets, are carried as clinical-pathology-style records in LB with supporting qualifiers in SUPPLB (assay, result type, titer, and impact flags). Reviewers should read ADA data in conjunction with the toxicokinetic (PC/PP) data, since the target-mediated disposition of this antibody means that exposure and ADA status must be interpreted together (see Section 8).

6. Controlled Terminology and Coding

Findings, specimens, anatomic locations, and test names use CDISC SEND Controlled Terminology at the package version recorded in the define.xml and TS domain. Sponsor-defined terms were introduced only where no CDISC term existed; each is documented in the define.xml (codelist/value-level metadata) with its extension flagged. Histopathology (MI) severity is recorded using the standard controlled grading scale, with the original pathologist's terminology preserved in the original-result variables. Because nonclinical SEND does not employ MedDRA or WHO Drug, no external medical or drug dictionary versions apply to these datasets; the only dictionary component is the CDISC SEND CT package identified above, applied uniformly across the four studies.

7. Study Design Representation (Trial Design Domains)

For each study the trial-design domains reconstruct the design a reviewer sees in the study-report methods section. TA encodes four arms for the general-toxicology studies — vehicle control and 10, 30, and 100 mg/kg — and the two treated arms plus control for the ePPND study. TE separates the acclimation (pre-treatment) element, the once-weekly SC treatment element (13 or 26 weeks for the chronic studies; 4 weeks for TILA278-NC-101), and the treatment-free recovery element. TX (Trial Sets) captures the main-study versus recovery-subset structure and sex, so that the reviewer can confirm that recovery animals were assigned only to the control and high-dose arms as designed. SE (Subject Elements) then shows the elements each animal actually experienced, which is the basis for distinguishing scheduled terminal-sacrifice animals from recovery-sacrifice animals in DS. TS records the study-level constants — species (cynomolgus monkey), route (SC), dosing frequency (once weekly), study duration, NOAEL as stated in the report, and the standards/CT versions — providing the machine-readable summary that anchors the rest of the datasets.

8. Toxicokinetics, Immunogenicity, and Specialized Data

Toxicokinetic sampling times, serum TILA-278 concentrations, and derived parameters (Cmax, Tmax, AUC over the dosing interval, and accumulation across weeks) are represented in PC and PP, with RELREC linking each concentration profile to its EX dosing event. Two properties of this molecule should guide interpretation of the TK data:

• Target-mediated drug disposition (TMDD). TILA-278 binds soluble/membrane TL1A and cell-surface IL-22R, and elimination includes a saturable, target-mediated component in addition to linear catabolic clearance. The result is nonlinear (dose-dependent) exposure at lower dose levels and more nearly dose-proportional exposure once target is saturated at the higher levels. The PP parameters should therefore not be assumed to scale linearly across the 10/30/100 mg/kg range; the study reports present the exposure–dose relationship explicitly.
• Immunogenicity. As a humanized antibody dosed in a non-human primate, TILA-278 can elicit anti-drug antibodies. ADA can accelerate clearance and depress exposure in individual animals, particularly at later timepoints and at the lowest dose. When reconciling an animal's TK profile against the group summary, reviewers should check that animal's ADA status; the LB/SUPPLB ADA records and the PC/PP records are intended to be read together, and RELREC does not force this link, so it is called out here.

No stand-alone safety-pharmacology datasets are provided because those endpoints (cardiovascular telemetry/ECG, respiratory, and neurobehavioral assessments) were integrated into the repeat-dose studies and appear within the CV/EG, RE, NV, and VS findings of TILA278-NC-101/102/103, consistent with ICH S6(R1) and ICH S7A. As noted in Section 1, there are deliberately no genetic-toxicology, carcinogenicity, hERG, or thorough-QT datasets for this antibody.

9. Data Conformance Summary

Each study's datasets were validated with Pinnacle 21 against the SENDIG v3.1 rule set, and the define.xml was validated against the Define-XML v2.0 schema. The conformance reports are provided with the submission. No errors that would impair loading or review remain unresolved. The residual messages are limited to warnings and notices that are expected for this program and are explained rather than corrected, including:

• Controlled-terminology extension notices arising from sponsor-defined terms that have no CDISC SEND CT equivalent (documented in define.xml with the extension flagged).
• Expected "domain not present" notices for domains that are legitimately absent for a monoclonal antibody program (no genetic-toxicology, carcinogenicity, tumor-findings (TF), or reproductive endpoints in the general-toxicology studies).
• Records intentionally flagged as not-done or excluded where a scheduled collection was not performed for a documented in-life reason, with the reason preserved in the status/reason variables.

A dispositioned message-by-message table is included with the Pinnacle 21 output; no rule failure changes any reported result or affects the NOAEL determinations described in Module 2.6.

10. Define.xml and Submission Package Organization

The datasets are organized under the eCTD Module 4 datasets folder, one subfolder per study (TILA278-NC-101, TILA278-NC-102, TILA278-NC-103, TILA278-NC-201). Each study folder contains its XPT datasets, its study-specific define.xml v2.0, the define2-0-0 rendering stylesheet, and the study's Pinnacle 21 conformance report. This nSDRG (SEND-001) sits at the datasets level and applies across all four studies. The define.xml is the entry point for machine-assisted review: it enumerates the datasets present for each study, provides variable- and value-level metadata, documents every controlled-terminology codelist and each sponsor extension, and records the SENDIG, Define-XML, and CDISC SEND CT versions. Opening the define.xml with the supplied stylesheet reproduces the domain inventory of Section 5 for each study.

11. Reviewer Notes and Known Issues

• One species by design. The cynomolgus monkey is the only pharmacologically relevant species for TILA-278; the absence of a rodent general-toxicology dataset is scientifically justified in the Nonclinical Overview and is not a data gap.
• Deliberate domain absences. No genetic-toxicology, carcinogenicity, tumor-findings, hERG, or thorough-QT datasets are present, consistent with ICH S6(R1) for a monoclonal antibody. Safety-pharmacology endpoints are integrated within the repeat-dose studies rather than submitted separately.
• Read TK, ADA, and dose level together. Because of target-mediated disposition and potential anti-drug antibody formation, exposure is nonlinear across the 10/30/100 mg/kg range and can be depressed in individual ADA-positive animals; PC/PP, LB/SUPPLB, and EX should be reconciled jointly.
• Verbatim preservation. Where CDISC SEND CT lacked a term, the original collected text is retained in the original-result variables and the sponsor extension is documented in define.xml; no collected value was altered in conversion.
• Authoritative source. For any discrepancy in interpretation, the GLP final study reports in Module 4.2 and the summaries in Module 2.6 govern; the SEND datasets are a faithful standardized representation of those reports.