BD teams evaluating China-origin assets understand that NMPA approval is not FDA or EMA approval. What gets underestimated, consistently, is how much work sits between the two. It’s operational, documented, and expensive. Getting it wrong at term sheet stage means inheriting a program that will cost twice what the model said and take years longer than the timeline projected.
Why the Gap Exists
China joined ICH in June 2017 and has formally adopted 59 of 63 ICH guidelines, including Q8 (Pharmaceutical Development), Q9 (Quality Risk Management), and Q10 (Pharmaceutical Quality System). On paper, the regulatory alignment is real and has been building for nearly a decade.
In practice, formal adoption and actual implementation at a specific manufacturing site are two different questions. NMPA inspection findings continue to surface serious CMC deficiencies at domestic manufacturers, including failures to demonstrate quality consistency with reference drugs and API sourcing from non-validated suppliers. Second and third-tier manufacturers have been the subject of China’s own quality consistency reevaluation program, launched in 2016 and extended to injectables in 2020, precisely because many generic products failed to meet comparator standards under closer scrutiny.
There is also a structural difference in what the regulators ask for clinically. NMPA has historically weighted endpoints and comparators relevant to the Chinese patient population. FDA and EMA require bridging to their own populations, against the current standard of care in those markets, not the comparator arm that was appropriate in China at the time of the pivotal trial.
The Clinical Data Gap
China-conducted trials often enroll homogeneous populations and use comparators that reflect Chinese standard of care at the time of study design. Neither of those things makes the data invalid. Both of them create questions that FDA and EMA will ask explicitly.
The sintilimab review at ODAC in March 2022 is the most cited example. The committee voted 14 to 1 that additional US-applicable trials were required before a final approval decision could be made. The pivotal ORIENT-11 trial had used chemotherapy alone as the comparator. At the time of the review, US standard of care for the relevant indication was pembrolizumab plus chemotherapy. The committee’s concern was not that the data was unreliable. It was that the study was not designed to answer the question FDA needed answered for the US population.
For deal teams: if the pivotal trial uses a comparator that is no longer standard of care in the target market, the cost of an additional bridging or confirmatory trial needs to be in the model before heads of terms are signed.
CMC and Manufacturing
This is where the gap is widest and most consistently underpriced at deal stage.
FDA inspection data for Chinese manufacturing sites shows a combined VAI and OAI rate of 69.0% across FY2014 to FY2024, based on a 2026 peer-reviewed analysis published in the Journal of Pharmaceutical Innovation. In May 2025, FDA expanded its policy on unannounced foreign inspections, increasing the frequency of surprise inspections at sites outside the US, including China.
Common CMC gaps that surface during FDA and EMA review include:
- Process validation packages that lack adequate process performance qualification data or rely on too few validation batches to support commercial scale conclusions
- Analytical method validation that meets NMPA standards but falls short of ICH Q2(R1) requirements for specificity, robustness, or intermediate precision
- Stability data generated under Zone II conditions (25°C/60% RH) without Zone IVa data (30°C/65% RH) where tropical market filings are intended, or Zone IVb (30°C/75% RH) where Article 58 or non-EU market filings are required
- Impurity profiling that has not addressed ICH M7 thresholds for mutagenic impurities
- API sourcing from suppliers that have not undergone Western-standard qualification audits and lack the traceability documentation FDA or EMA will look for
None of these are fatal. All of them take time and money the deal economics need to account for before close.
Data Integrity
FDA and EMA inspect for data integrity at a level of rigor that NMPA has historically not required to the same specification. Chinese sites that have never been inspected by a Western authority often lack the documentation infrastructure FDA and EMA expect to see. This is not a statement about intent. It is a statement about infrastructure.
What FDA and EMA look for during a data integrity inspection:
- Electronic audit trails that are complete, unedited, and retained at the raw data level
- Access controls that prevent backdating or unauthorized data entry
- 21 CFR Part 11 compliance (FDA) or EU Annex 11 compliance (EMA) for computerised systems
- Original source documents for all GMP and GCP data, not just summarised reports
For clinical data specifically, FDA and EMA apply ALCOA+ standards: records must be Attributable, Legible, Contemporaneous, Original, and Accurate, plus Complete, Consistent, Enduring, and Available. A review of source documents, not just the summarised data package, is required before any Western regulatory submission relies on the China-generated dataset.
Rebuilding data integrity infrastructure after an inspection finding, under remediation conditions with a regulator watching, is significantly more expensive than building it correctly before the NDA or MAA is filed.
Stability Data
Stability zone requirements are a frequent source of confusion when evaluating China-origin assets. NMPA submissions often include data generated under conditions appropriate for Chinese or Asian markets that do not map directly to EU MAA requirements. The table below covers what each zone means and when it applies:
| Zone | Conditions | Applies to |
|---|---|---|
| Zone II | 25°C / 60% RH | Standard EU MAA (long-term) |
| Zone IVa | 30°C / 65% RH | Alternative long-term; some tropical markets |
| Zone IVb | 30°C / 75% RH | Article 58 filings; non-EU market registrations only |
| Accelerated | 40°C / 75% RH | Supports initial dossier submission; not a substitute for long-term data |
If Zone II long-term data does not exist for an asset, generating it takes a minimum of 12 months. That timeline is fixed by the ICH Q1A(R2) protocol and cannot be compressed, regardless of what accelerated data is available.
The BIOSECURE Act: What It Actually Covers
The BIOSECURE Act was enacted on December 18, 2025, as Section 851 of the FY2026 National Defense Authorization Act (P.L. 119-60). It restricts federal procurement, contracts, grants, and loans involving entities designated as “biotechnology companies of concern” (BCCs). No companies are named as BCCs in the enacted statute. That designation comes through an OMB-led process, with the initial BCC list due by December 18, 2026.
What the Act does not do: it does not prohibit private-sector in-licensing of Chinese-origin pharmaceutical assets. It does not automatically restrict CDMO or CRO relationships with Chinese service providers unless there is a federal contract, grant, or loan nexus. The flow-down risk is real for federal contractors whose supply chains touch a designated BCC, but the exposure for a purely private, non-federally funded licensing transaction depends on which companies end up on the BCC list and what the relevant supply chain looks like.
BD teams working with Chinese CDMO or CMO supply chains should track the OMB designation process through 2026. The current statute is not the constraint. The uncertainty around the forthcoming list, and its potential impact on supply agreements already in place, is.
Building a Realistic Bridging Plan
The programs that bridge successfully are not necessarily the ones with the cleanest data going in. They are the ones where the Western partner understood the bridging requirements before signing and built the cost and timeline into the deal structure.
A realistic bridging plan needs to address:
- Gap analysis before close: an independent CMC and clinical review against FDA and EMA standards, conducted before heads of terms are signed rather than after
- Manufacturing remediation or transfer: timeline and cost to bring the site to Western GMP standard, or a plan to transfer manufacturing to a qualified CMO with a Western inspection track record
- Clinical bridging strategy: whether existing data can support a 505(b)(2) pathway in the US, or whether new US or EU-applicable trials are required. EU hybrid MAA applications carry a substantially higher evidentiary bar and should not be assumed as a straightforward parallel to 505(b)(2)
- Stability program: whether compliant Zone II long-term data exists for EU MAA, and if not, when it can be generated and how that affects the filing timeline
- Data integrity audit: source data review for both GCP and GMP data before any Western regulatory submission relies on the dataset
- Regulatory alignment: FDA pre-IND or Type B meeting, EMA scientific advice, agreed and documented before the bridging program clock starts running
No systematic published data exists on the average timeline from NMPA approval to FDA or EMA approval for China-origin assets requiring bridging. The sample of completed cases is too small and too recent. The available case examples suggest a range of 3 to 6 years: toripalimab took approximately 2.5 to 5 years across its indications; tislelizumab approximately 3 to 4 years. Both required additional engagement with FDA on the representativeness of China-conducted trials for the US population. For financial modelling, the 3 to 6 year range is the honest anchor, with the lower end applying only where a coherent bridging strategy was in place before the deal closed.
For context on how China-origin assets actually reach EU and US markets and what the deal economics look like, see our earlier analysis on this topic.