What makes a trial landscape search hard

A single mechanism-and-disease question rarely has a single-database answer. Registered trials are split across national and international registries with their own field names and update schedules, and a registration record doesn’t always link cleanly to its published results — those can sit in a separate results database entirely. Getting a real competitive picture means combining several filters at once — mechanism, disease, sponsor, phase, geography — rather than a single keyword, and then cross-referencing registration entries against outcome data by hand.

Why this needs Clinical Trials

An AI assistant or coding agent can explain trial phases and suggest search terms from its training data, but it can’t retrieve a live registry record, filter by sponsor or disease, or tell you how many matching studies exist today — that answer stays hypothetical without a live connection. Clinical Trials adds structured search across drug, mechanism of action, disease, sponsor, phase, and region in one query, plus a separate trial-result search so registration and outcome data live in the same session instead of two disconnected lookups.

Prepare your query

Start with the clinical or competitive question you want answered — which trials test a given mechanism in a specific indication, or what Phase 3 programs a competitor is running — then identify the filters that scope it. This tutorial’s question is “What are the registered clinical trials for PD-1 inhibitors in colorectal cancer?”, using mechanism of action “PD-1 inhibitor” and disease “colorectal cancer” as the two filter terms.

1Connect Clinical Trials

1. Get a Patsnap Open Platform API key.
2. Copy the connection URL or JSON config from the Connect panel on the Clinical Trials page.
3. Add the server to your AI assistant or coding agent or another MCP-compatible client, then restart it.
4. Confirm the trial search, result search, and batch fetch tools are available before you ask.

2Run the example query

Find clinical trials for PD-1 inhibitors in colorectal cancer.

Claude recognizes the request and calls clinical_trial_search(mechanism_action=["PD-1 inhibitor"], disease=["colorectal cancer"]). The server applies both filters and returns matching trials with registration numbers, titles, phases, interventions, and sponsor details.

3What the result tells you

Nearly a thousand registered records link PD-1 checkpoint inhibition to colorectal cancer research, and the top result surfaces the most relevant program: at least one late-stage trial is evaluating a PD-1 agent combined with an anti-angiogenic antibody in this indication. Each returned record is a registered study — use its registry number to pull the full record or track updates in a follow-up query.

4Go further with follow-up questions

From the same session, narrow or pivot with:

• “Which of these trials are Phase 3 and sponsored by Chinese organizations?” — adds phase and sponsor filters for late-stage regional competitive programs.
• “Show me trials combining PD-1 inhibitors with chemotherapy in colorectal cancer.” — refines the intervention to combination regimens.
• “Retrieve the full details for the IBI363 trial using its registry number.” — uses the batch fetch tool for a complete record.