How to Map a Clinical Trial Landscape for a Drug Target

If you're an agent: Set up Valyu and make your first search.

If you're a human: Create an account and get an API key.

MCP onboarding: Connect your agent to https://mcp.valyu.ai/mcp and follow the MCP setup guide for authentication.

What will the landscape show?

Suppose you work in biotech business development and a colleague asks whether Claudin 18.2 is still open for a new entrant in gastric cancer. You need to know who is running what, and how far each program has got. You ask:

Which phase 3 trials target Claudin 18.2, who sponsors them, and what has actually been disclosed?

The useful output is a single table that ties each trial to its registry record and any dated results, publication or approval. Then you can see which programs have evidence and which are only registered.

Step 1: define the target and the fields

Decide what belongs on the map before you search:

• Target and synonyms: Claudin 18.2, CLDN18.2 and known drug names.
• Scope: the phase and whether you want all indications. Our example covers phase 3 across indications.
• Inclusion rule: the trial must test a therapy aimed at the target. A trial that only mentions the target in its eligibility criteria doesn't count.
• Fields: NCT ID, sponsor, drug, modality, indication, line of therapy, status, enrollment and dated disclosures.

Step 2: run the map with Valyu

One of the ways to run it is with Valyu connected to Claude or ChatGPT. Follow the chat apps guide to connect it, then ask:

Use Valyu DeepResearch in fast mode to map the phase 3 clinical trial landscape for Claudin 18.2 targeted therapies. For each trial give the NCT ID, sponsor, drug and modality, indication, line of therapy, phase, registry status, enrollment, and any dated result disclosure with its source. Separate registry status from published or presented results. Note regulatory approvals with dates and sources. Do not infer results from registry status. List gaps and mark anything unverified. Ask me to confirm before you start.

You can also run it from your own application with the SDK. Install it with pip install valyu for Python or npm install valyu-js for TypeScript, and set VALYU_API_KEY in your server environment. The Python SDK and TypeScript SDK docs cover the options. Both examples send the same request:

from valyu import Valyu
 
client = Valyu()
 
task = client.deepresearch.create(
    query=(
        "Map the phase 3 clinical trial landscape for Claudin 18.2 targeted therapies. "
        "For each trial give the NCT ID, sponsor, drug and modality, indication, line "
        "of therapy, phase, registry status, enrollment, and any dated result "
        "disclosure with its source. Separate registry status from published or "
        "presented results. Note regulatory approvals with dates and sources. Do not "
        "infer results from registry status. List gaps and mark anything unverified."
    ),
    mode="fast",
    research_strategy=(
        "Start from ClinicalTrials.gov records, then look for primary publications "
        "and regulator documents for each trial. Check each NCT ID against the drug "
        "it is attributed to."
    ),
    report_format=(
        "A concise brief with a trial table. Cite every finding and separate direct "
        "disclosure from interpretation."
    ),
)
if not task.success or not task.deepresearch_id:
    raise RuntimeError("No task ID returned; inspect the submission before retrying")
 
print("Save this task ID:", task.deepresearch_id)
result = client.deepresearch.wait(
    task.deepresearch_id, poll_interval=5, max_wait_time=900
)
if result.status != "completed":
    raise RuntimeError("Research has not completed")
 
print(result.output)

Check out all of the other integration options here: https://docs.valyu.ai

Step 3: read the map

Three trials have disclosed results. We confirmed these details in the registry, PubMed, the FDA's announcement and AstraZeneca's release:

Two approvals sit alongside these. The FDA approved zolbetuximab on October 18, 2024 for first-line use, based on SPOTLIGHT and GLOW. The report also cites a China approval for satri-cel, a CAR-T therapy, in June 2026. That rests on a phase 2 trial and a secondary source, so confirm it with the company or regulator before using it.

The other 17 phase 3 trials had no disclosed results in our check:

• Active, not recruiting: IBI343 (NCT06238843), LM-302 (NCT06351020) and ASKB589 (NCT06206733).
• Recruiting: zolbetuximab plus pembrolizumab (NCT06901531), sonesitatug vedotin first line (NCT07431281), ASP2138 (NCT07673887), SHR-A1904 (NCT06649292), JS107 (NCT07284134 and NCT07584135), LM-302 plus tislelizumab (NCT07385703), QLS31905 (NCT07079228 and NCT07716449), BL-M05D1 (NCT07518147), ATG-022 (NCT07795346) and IMC002 (NCT07103668).
• Not yet recruiting: osemitamab (NCT06093425) and FG-M108 (NCT07383922).

Step 4: cross-check the list against the registry

A landscape is only useful if it is complete, so don't accept the first list. We searched ClinicalTrials.gov directly with target synonyms and the phase filter, then compared the results with the report:

• The report found 14 trials. It said its coverage wasn't exhaustive, and it missed six that we found: LM-302 monotherapy, the second JS107 trial, ASP2138, ATG-022, ASKB589 and FG-M108.
• Keyword search missed two. Our registry query didn't return IBI343 or BL-M05D1, even though the report had them and both records mention the target.
• One match wasn't on target. The search also returned a trifluridine/tipiracil plus fruquintinib trial (NCT07270991), which mentions Claudin 18.2 in its background and eligibility text but doesn't test a therapy aimed at it. We left it out.

Together that makes 20 trials. The report's prose also mentions a bispecific among the modalities, while its own gaps section says several modalities were unverified. So check modality against a company or scientific source before you group programs by it. Search again by each drug name too, because that's the fastest way to catch a sponsor's second trial.

Step 5: check every result at its source

Headline numbers often describe a specific population. AstraZeneca's release for CLARITY-Gastric01 shows why you should read past the headline:

• Second line and later (overall population): median overall survival of 10.6 versus 6.4 months, HR 0.58 (95% CI 0.47–0.73). This is the 42% figure in the headline.
• Third line and later (a primary endpoint): median overall survival of 10.0 versus 5.0 months, HR 0.47 (95% CI 0.36–0.62).
• Progression-free survival (the other primary endpoint): HR 0.77 (95% CI 0.62–0.96), p=0.053193, which did not reach statistical significance.

The report kept these apart correctly, including the missed endpoint. Also note the source type: this is a company release of conference data, not a peer-reviewed paper, so label it that way on your map.

Figure 1. Three programs with the same registry status and very different evidence, as of November 5, 2026.

How should you interpret the map?

The decision: treat first-line gastric cancer as a market with an approved incumbent, zolbetuximab, and several late-stage challengers. Treat later lines as promising but contested, with a positive survival readout, a missed co-primary endpoint and multiple competing trials.

The next action: confirm modality and line of therapy for every program on the map, check other national registries where sponsors may also list trials, and add the expected readouts. For example, IBI343 and LM-302 monotherapy have primary completion dates estimated for late 2026.

The outcome: a dated, one-page map your team can defend, with registry facts and evidence kept apart, and a watchlist of readouts to review each quarter.

How to adapt the map

Swap the target and phase, and keep the rule that status and results stay separate. For a repeatable run with a spreadsheet, the trial landscape guide shows how to request a CSV. To find trials a patient could join, see finding recruiting trials via API. For acquisition work, see finding acquisition targets from a thesis.

FAQ

What is a clinical trial landscape?

It's a structured view of the trials on a target, drug class or disease, showing sponsors, drugs, phases, status and results. Teams use it for competitive analysis, licensing decisions and trial design.

Why isn't registry status enough?

Status describes operations such as enrollment. SPOTLIGHT is still listed as active, even though it was published in 2023 and supported an FDA approval in 2024.

How do I make sure I haven't missed a trial?

You can't be certain, so reduce the risk. Combine a research report with direct registry searches, use target synonyms, search each drug name, and check registries beyond ClinicalTrials.gov. Date your map so readers know when it was last checked.

Can I trust a press release as a source?

Use it as a dated company disclosure, and label it that way. Replace it with the peer-reviewed paper when one appears, and check that the numbers still match.

Do I need a ClinicalTrials.gov API integration?

Not for this job. The Valyu connector searches trial records alongside papers and the web from a chat prompt. The official API is still the better choice for an exhaustive export of records.

Which DeepResearch mode should I use?

Fast mode gave us a strong first map, but it stopped at its step limit with six trials missing. For a crowded target, try a deeper mode and check the pricing first.

Build a landscape your team can defend

Define the target, run the map, cross-check it against the registry and read each result at its source. Give your agents access to Valyu, read the Valyu docs, explore the biotech and pharma documentation, or read about building a biomedical research agent.