For the complete documentation index, see llms.txt. This page is also available as Markdown.

Navigate clinical trials

What is Trial Navigator?

Trial Navigator is a specialized agent in Owkin K that helps pharmaceutical and clinical teams explore the clinical trial landscape using plain-language questions. You ask the way you'd ask a colleague; it interprets the intent, filters a large clinical-trials dataset, and returns the right output: a visual timeline, a data table, a statistical projection, or a direct answer.

Use it to:

  • Map competitive landscapes across an indication, mechanism of action, or drug class

  • Track competitor pipelines and development timelines

  • Answer quantitative questions about the trial landscape (counts, averages, distributions)

  • Compare trials against a reference trial (e.g. "PFS better than trial X")

  • Deep-dive a single trial's design assumptions and projected readout times

How it works

You don't need to know the internals to use Trial Navigator, but a quick mental model helps you write better queries. Behind the scenes it:

1

Plans

Breaks a complex question into sub-questions and, when needed, runs them in sequence, so "find trials better than X" first finds X, then compares the rest of the landscape against it.

2

Translates

Turns your words into structured filters: indication, phase, sponsor, dates, mechanism of action, enrollment thresholds, and more.

3

Reconciles terminology

Maps your wording to the dataset's standard names via a built-in thesaurus, so "immunotherapy", "IO", and "PD-1 inhibitor" resolve to the right underlying terms.

4

Answers

Returns a Gantt chart and table, a statistical projection, and/or a written summary, depending on what you asked for.

You can be conversational and use common abbreviations, the tool is designed to resolve them. The more specific you are, the sharper the result.

See it in action

Writing effective queries

  • Name the therapeutic area / indication (e.g. NSCLC, TNBC, glioblastoma).

  • Name the mechanism of action or drug class when relevant (PD-1/PD-L1, EGFR TKI, ADC, VEGFR, CLDN18.2).

  • Name sponsors, drugs, or NCT IDs when you have them.

  • Add a phase to narrow the development stage (Phase I, II, III, I/II…).

  • Add a timeframe ("started after 2020", "between 2018 and 2023").

  • Add thresholds ("enrollment greater than 500", "accrual target of exactly 500").

  • Use exclusions freely ("excluding Phase I", "excluding breast and prostate cancer").

  • Ask for top / latest / longest when you want a ranked shortlist ("the 5 most recent…", "longest-running…").

What you can ask

Competitive landscape / list trials

Returns an interactive Gantt chart, a detailed results table, and summary statistics.

Complex multi-filter queries

The tool handles several constraints at once, combining indication, phase, geography, dates, mechanism, and thresholds.

Ranked / "top N" / "latest N" / "longest-running"

Ask for a ranked shortlist instead of the full set.

Aggregate & quantitative questions

For counts, averages, percentages, and "what's covered" questions, the tool answers directly in text, with supporting figures where useful.

Comparative queries (against a reference trial)

The tool can find a reference trial first, then compare the rest of the landscape against it.

Single-trial statistical analysis

Deep-dive the statistical assumptions behind a specific trial's design and projected readout times. Returns a projection plot, predicted median OS/PFS, predicted enrollment duration, and the reference trials used for the prediction.

Understanding the outputs

Gantt chart (timeline)

  • X-axis: time (years)

  • Y-axis: individual trials, grouped by phase and sorted by start date

  • Colour: trial phase; bars show duration with mechanism-of-action abbreviations

  • Includes both industry-sponsored and publicly/academically sponsored trials.

Results table

A sortable, filterable list of every matching trial: NCT ID (linked to ClinicalTrials.gov), sponsor, phase, indication, drug/mechanism, and key results.

Summary statistics

An automatically computed answer summary, for example total trial count, phase distribution, and top indications represented.

Statistical projection plot

For single-trial analysis: an enrollment curve and survival projections, with the reference trials (by NCT ID) used to generate the predictions.

PowerPoint export

Trial Navigator can export a competitive-landscape view into a PowerPoint slide.

PowerPoint export is a newer, deployment-gated capability, availability depends on your K environment.

Searchable data fields

Trial Navigator is built on a comprehensive oncology clinical-trials dataset sourced from ClinicalTrials.gov. You can filter and search across the following dimensions.

Identification

Field
Description

NCT ID

ClinicalTrials.gov identifier

Brief Title

Short trial title

Official Title

Full formal trial title

Commercial Name

Sponsor's commercial trial name

Status & timeline

Field
Description

Overall Status

Recruiting, Active Not Recruiting, Completed, Terminated, etc.

Trial Start Date

Date the trial began

Primary Completion Date

Date of primary outcome measurement

Completion Date

Full study completion date

Trial Duration

Duration in days

Design

Field
Description

Study Type

Interventional, Observational

Trial Phase

I, II, III, IV, I/II, II/III, III/IV

Randomization

Whether the trial is randomized

Number of Arms

Count of treatment arms

Group Types

Experimental, Active Comparator, Placebo Comparator, etc.

Sponsors

Field
Description

Private / Industry Sponsors

Pharmaceutical and biotech companies (Pfizer, Roche, AstraZeneca, etc.)

Public / Academic Sponsors

Universities, hospitals, research centers, government agencies

Industry Sponsored

Boolean flag for industry involvement

Enrollment & accrual

Field
Description

Enrollment

Number of participants

Accrual Target

Target number of subjects

Accrual Actual

Actual enrolled subjects

Enrollment Duration

Time to complete enrollment (days)

Geography

Field
Description

Trial Countries

Countries where the trial is conducted

Trial Regions

Regions (North America, Western Europe, Asia, etc.)

Number of Sites

Count of trial sites

Eligibility criteria

Field
Description

Gender

Both, Female Only, Male Only

Age Range

Minimum and maximum age for eligibility

ECOG Status

Performance status requirements (0-4 scale)

Disease Stages

Localized, locally advanced, advanced, metastatic, resectable, unresectable

Disease Paths

Persistent, recurrent, refractory, relapsed, resistant

Lines of Therapy

Prior treatment requirements

Indications (cancer types)

Trial Navigator covers a comprehensive oncology landscape, including:

Solid tumors:

  • Breast, Lung (NSCLC, SCLC), Colorectal, Pancreas

  • Melanoma, Renal, Bladder, Prostate

  • Head/Neck, Esophageal, Gastric, Liver

  • Ovarian, Endometrial, Cervical

  • CNS tumors (Glioblastoma, Astrocytoma, Meningioma, etc.)

  • And many more

Hematological malignancies:

  • Leukemia (ALL, AML, CML, CLL)

  • Lymphoma (Hodgkin's, Non-Hodgkin's)

  • Multiple Myeloma

  • Myelodysplastic Syndrome

  • Myeloproliferative Neoplasms

Treatment & intervention

Field
Description

Drug Names

Investigational drug names

Drug Type

ADC, Antibody Bispecific, Antibody Multispecific, Gene Therapy, Small Molecule

Mechanisms of Action

JAK inhibitor, CDK inhibitor, PD-1/PD-L1 inhibitor, etc.

Targets (HGNC)

Gene symbols (KRAS, TP53, BRCA1, EGFR, ERBB2, etc.)

Biomarkers

Biomarkers used in trial selection

Combination Therapy

Whether the trial tests a drug combination

Endpoints

Field
Description

Primary Endpoints

OS, PFS, DUAL, or other

Has OS

Overall Survival as endpoint

Has PFS

Progression-Free Survival as endpoint

Has EFS/DFS/RFS

Event-free, Disease-free, Recurrence-free Survival

Outcomes (actual & predicted)

Field
Description

Objective Response Rate

ORR from trial results

Disease Control Rate

DCR from trial results

Median OS (months)

Actual median overall survival

Median PFS (months)

Actual median progression-free survival

Predicted Median OS

AI-predicted median OS

Predicted Median PFS

AI-predicted median PFS

Predicted Enrollment Duration

AI-predicted time to complete enrollment

Tips & good-to-knows

  • Be conversational, but specific. Abbreviations and synonyms are resolved by the thesaurus; specificity narrows results.

  • Smart name resolution. The tool understands common synonyms and variations, for example "Keytruda" maps to pembrolizumab, "Herceptin" to trastuzumab, gene-name variants (ERBB2 / HER2 / neu), and sponsor name variations (Merck & Co., MSD, Merck Sharp & Dohme).

  • Large landscapes are shown, not silently truncated. The chart focuses on the most relevant trials for readability while the underlying query covers the full matching set. If a very large landscape is capped for display, the tool tells you.

  • Statistical projections are model-based estimates, derived from historically comparable reference trials. Treat them as directional, not definitive, and note they are available only for Phase II and III trials with sufficient historical data.

  • Coverage is oncology-focused. Non-oncology questions may return limited or no results.

  • Data reflects ClinicalTrials.gov, so completeness depends on what sponsors have registered and published. For real-time status, verify directly on ClinicalTrials.gov.

Last updated

Was this helpful?