Dual‑track trials. Keep China speed, global options.

Clinical Trial Design: Keeping China's Speed While Building Global Datasets

China's enrollment velocity is a major advantage for early clinical work—you can hit recruitment targets in months that would take years elsewhere.

But a China-only dataset is increasingly a liability for US/EU partnerships or exits.

The question: How do you keep China's speed while building data that travels?

The Pattern: China-Only Data Becomes a Partnership Issue

What this looks like:

Your Phase 2 trial enrolled beautifully in China—faster than projections, good data quality, hit your endpoints.

18 months later, you're in partnership discussions with a US pharma. Their regulatory team asks:

• "Can we file in the US with this China-only dataset?"

• "Are the endpoints aligned with FDA expectations?"

• "Do we have enough ex-China patients to support a bridging claim?"

If the answers are unclear, the partnership conversation gets complicated—not because your data is bad, but because it's not portable.

Is this your situation?

If you're running or planning trials in China, ask your team:

• Are our endpoints harmonized for FDA/EMA acceptance?

• Do we have ex-China sites planned, or is this China-only?

• Can we pool data across regions, or are protocols too different?

If you're optimizing purely for enrollment speed without thinking about regulatory portability, you might be creating a filing bottleneck.

What's Working: Dual-Track Design from the Start

The teams getting both speed AND portability design their trials for dual-track enrollment from the beginning—China for velocity, ex-China for regulatory breadth.

What "dual-track" actually means:

Not running two separate trials. Running one trial with:

• Shared protocol and endpoints

• Parallel enrollment in China and ex-China sites

• Harmonized data collection and analysis plan

Three design choices that matter most:

1. Harmonize endpoints upfront (not after)

The most common mistake: designing a trial for China enrollment, then trying to "bridge" to US/EU regulatory requirements later.

What this creates: Endpoint definitions that don't quite match FDA/EMA expectations, assays that aren't comparable across regions, timing windows that differ enough to make pooling questionable.

What's working:

Design endpoints that work for NMPA (China), FDA (US), and EMA (EU) from day one.

What this looks like:

• Primary endpoint: Use definitions that all three agencies accept (e.g., progression-free survival with RECIST criteria, not regional variations)

• Biomarkers: Validate assays at labs in both China and ex-China, document comparability

• Visit windows: Keep timing consistent across regions (same blood draws, same imaging schedules)

• Statistical analysis plan (SAP): Pre-specify how you'll pool or compare data across regions

Why this matters:

If you try to harmonize after enrollment starts, you're stuck. The data you've already collected can't be retroactively aligned.

One company's experience:

A rare disease biotech designed their Phase 2 trial to enroll primarily in China (faster) but opened 3 US sites and 2 EU sites in parallel from the start.

The enrollment was still 70% China (kept the speed advantage), but having 30% ex-China patients meant:

• They could pool data for global filing

• They had safety/efficacy data in Western populations

• Partnership discussions didn't stall on "can you file in the US?"

The ex-China sites added ~$800K and 3 months to startup, but saved them from needing a full bridging study later (which would have been $5M+ and 18+ months).

2. Activate ex-China sites early (not when you realize you need them)

The second most common mistake: enrolling only in China initially, then trying to add ex-China sites later when partnership discussions surface the need.

What this creates:

• Ex-China sites start 12-18 months behind China sites

• By the time they're enrolling, your China sites are finishing

• You miss the window for concurrent data collection

• You need a sequential amendment or bridging study

What's working:

Open a small number of ex-China sites at the same time as (or shortly after) China sites—even if most enrollment will be in China.

What this looks like:

Target enrollment mix examples:

• Phase 1: 80% China / 20% ex-China (keeps speed, adds safety data diversity)

• Phase 2: 60-70% China / 30-40% ex-China (balanced for global filing)

• Phase 3: Depends on indication, but usually closer to 50/50 if global filing is the goal

The calculus:

Opening ex-China sites earlier costs more upfront (site startup, slower enrollment). But it's usually cheaper than:

• Running a separate bridging study (typical cost: $3-8M)

• Delayed partnership/exit due to data gaps (opportunity cost)

• Re-enrolling patients in ex-China to meet regulatory requirements

When this matters most: If your endgame is US/EU partnership or global filing. If you're targeting China-only or Asia-only markets, keeping it China-focused makes sense.

3. Plan for data residency and cross-border compliance

Clinical trial data crosses borders constantly—CROs, central labs, data management, biostatistics. But some jurisdictions have strict requirements about where patient data can be stored and processed.

What catches teams off guard:

• EU patient data (GDPR) has residency requirements

• Some pharma partners require US-only data hosting

• China has data export restrictions for certain data types

• These requirements weren't built into the original trial design or contracts

What's working:

Define data residency requirements upfront, by data class:

Patient identifiable information (PII/PHI): Usually must stay in region of origin (EU data in EU, China data in China, etc.)

De-identified clinical data: More flexible, but check partnership agreements (some partners require specific data centers)

Biomarker/omics data: Can have export control issues depending on technology and region

What this looks like in contracts:

• CRO contracts specify where data is hosted by type

• Central lab agreements include data residency clauses

• Cloud/database vendors are selected based on regional data centers

Why this matters:

If you discover data residency violations during partnership diligence, you might need to:

• Migrate data (expensive, time-consuming)

• Get retrospective consent (often impossible)

• Exclude certain data from filing (weakens your case)

Better to design compliance in from the start.

One Full Example

A Phase 2 metabolic disease company wanted to leverage China's enrollment speed but knew they'd need US/EU partnerships eventually.

What they designed:

Protocol: Harmonized endpoints acceptable to NMPA, FDA, EMA from day one

Sites: 12 China sites (primary enrollment) + 4 US sites + 3 EU sites (opened in parallel)

Target enrollment: 70% China, 30% ex-China (actual ended up 65%/35%)

Data: Contracted CROs and central labs with appropriate data residency controls by region

Extra cost: ~$1.2M and 4 months additional startup time vs. China-only

Payoff: 18 months later during partnership discussions:

• Could pool all data for global filing (no bridging study needed)

• Had safety/efficacy in Western populations

• Partner comfortable with data quality and regulatory strategy

Their BD team said the dual-track design was explicitly called out by the partner as reducing risk and accelerating timelines.

How to Actually Do This

Step 1: Decide if dual-track makes sense (before protocol finalization)

Ask:

• Is our endgame US/EU partnership or global filing?

• Can we absorb the extra cost and time for ex-China sites?

• Are our endpoints already aligned for FDA/EMA, or would we need changes anyway?

If yes to first two questions, dual-track usually makes sense.

Step 2: Design protocol with harmonized endpoints

Work with regulatory consultants to ensure:

• Primary/secondary endpoints are acceptable to NMPA, FDA, EMA

• Biomarker assays are validated in both regions

• Visit schedules and data collection are consistent

This adds time to protocol development (2-4 weeks) but is much cheaper than fixing it later.

Step 3: Open ex-China sites early (not later)

Identify 3-5 ex-China sites and start activation in parallel with (or shortly after) China sites.

Target: ex-China sites open within 3-6 months of China sites starting enrollment.

Step 4: Build data compliance into contracts

Before signing CRO, central lab, or data management contracts, specify:

• Where data will be hosted (by data class)

• Cross-border transfer restrictions

• Partner audit rights

What Good Looks Like

You'll know dual-track is working when:

• Partnership discussions don't stall on "can you file in the US?"

• Your SAP clearly shows how regional data will be pooled or compared

• Regulatory consultants confirm your data package works for global filing

• You're getting enrollment speed from China without sacrificing portability

Most reliable signal: When potential partners or acquirers review your trial design and say "this is already set up for global filing" rather than "how will you address the China-only data gap?"

When This Actually Matters

Dual-track design is worth it if:

• You're Phase 2+ with US/EU partnerships likely

• Your indication supports global filing

• You can absorb the extra cost (~$1-2M and 3-6 months)

It's probably not worth it if:

• You're targeting China-only or Asia-only markets

• You're very early stage (Phase 1 safety, proof-of-concept only)

• Your endpoints are so China-specific they wouldn't transfer anyway

What Are You Seeing?

Are you designing trials with global portability in mind, or optimizing purely for enrollment speed? And if partnerships are asking about ex-China data, what's the actual gap—endpoints, patient population, data residency, something else?

Worth comparing notes.

—Roop