ICH E17 (“General Principles for Planning and Design of Multi-Regional Clinical Trials”) is the International Council for Harmonisation guideline that tells sponsors how to design a single clinical trial protocol intended to generate data acceptable to regulatory authorities in more than one region at once, rather than running separate, duplicative trials region by region. It reached ICH Step 4 (final guideline) in November 2017. The US FDA published it as final guidance for industry, with a Federal Register notice of availability dated 19 July 2018; Japan’s regulatory authority adopted it in June 2018, and the EU adopted it in July 2018. For a sponsor or CRO planning a multi-regional clinical trial (MRCT), E17 is the reference document for how the trial’s design, endpoints, and statistical analysis plan need to hold together across every participating region simultaneously — not just satisfy each region’s requirements independently after the fact.
What a Multi-Regional Clinical Trial Is, and Why It Needs Its Own Guideline
A multi-regional clinical trial is a single trial, conducted under one protocol and one overall development plan, with sites in more than one geographic region, where the intent from the outset is to use the pooled (and where appropriate, region-specific) results to support regulatory submissions in each of those regions. This is distinct from running fully separate national or regional trials and later trying to bridge or reconcile their results after the fact.
MRCTs became the dominant model for late-phase drug development because they let a sponsor generate one data package usable for simultaneous or near-simultaneous submissions worldwide, shortening time to approval and avoiding the cost and redundancy of duplicate regional trials. But pooling data across regions with different populations, healthcare systems, standard-of-care practices, and regulatory expectations creates real methodological risk: without deliberate planning, a sponsor can end up with a trial that is statistically valid in aggregate but uninterpretable, or unconvincing to a regulator, when a specific region’s subset of data is examined on its own. ICH E17 exists to make that planning explicit and consistent across ICH regions, rather than leaving it to each sponsor’s own ad hoc practice.
How E17 Builds on ICH E5
E17 does not start from nothing. It builds directly on ICH‘s earlier E5 guideline, “Ethnic Factors in the Acceptability of Foreign Clinical Data” (1998), which introduced the framework of intrinsic factors (genetic and physiological characteristics such as age, sex, organ function, and genetic polymorphisms) and extrinsic factors (environmental and cultural variables such as diet, medical practice, regulatory requirements, and socioeconomic conditions) that can cause a treatment’s effect to vary from one population or region to another. E5 was written primarily to guide after-the-fact bridging — assessing whether foreign clinical data already collected in one region could be extrapolated to support approval in another. E17’s contribution is to move that same intrinsic/extrinsic-factor thinking to the front end of trial planning: instead of collecting data in one region and later asking whether it transfers, a trial designed under E17 principles builds the ability to evaluate cross-regional consistency into the protocol itself, from population definition through statistical analysis plan.
Core Planning Principles Under E17
E17 sets out a small number of design principles a sponsor is expected to address and document when planning an MRCT:
- A single primary analysis approach for hypothesis testing. The trial should be designed around one pre-specified primary statistical analysis that applies across all participating regions, rather than each region effectively running its own parallel analysis on a shared dataset.
- Common comparators across regions. Where regional standard of care differs, the sponsor needs a documented rationale for the comparator(s) used and how that choice preserves the trial’s ability to answer its scientific question consistently everywhere it is conducted.
- A primary endpoint that is clinically relevant in every participating region. An endpoint that is meaningful and measurable in one region’s clinical practice but not in another undermines the entire premise of pooling the data.
- Precisely defined, cross-regionally applicable inclusion and exclusion criteria. Eligibility criteria have to be specific enough to be applied consistently by investigators working in different healthcare systems and regulatory environments.
- Uniform GCP compliance across every site, in every region. An MRCT is only as strong as its weakest region’s adherence to Good Clinical Practice — see ICH E6(R2) and its successor ICH E6(R3) for the conduct standard itself.
- Pre-specified monitoring plans and consistent sponsor-to-site information flow. Because an MRCT typically runs through multiple national coordinating investigators or regional CROs, the guideline expects a documented plan for how safety and operational information moves between sites, regions, and the sponsor without regional silos forming.
- Early engagement with regulatory authorities. E17 explicitly encourages sponsors to seek scientific advice or consultation meetings with the regulatory authorities of the regions they intend to include, early in trial planning — before the design decisions above are locked in — rather than discovering a region-specific objection only at submission.
Statistical Considerations: Evaluating Consistency of Treatment Effect Across Regions
A large part of E17’s substance concerns how a sponsor should plan, in advance, to evaluate whether the treatment effect observed is reasonably consistent across the trial’s participating regions — and how much regional variation is expected and acceptable versus a signal worth further investigation. This includes planning for the statistical allocation of sample size across regions (ensuring no single region is either too small to interpret or so dominant that it effectively determines the overall result on its own), and pre-specifying, in the statistical analysis plan, how consistency of effect will actually be assessed once data are unblinded — rather than deciding on an approach retrospectively based on how the data happened to come out. The guideline deliberately does not mandate one single statistical method for this evaluation; it sets out the planning expectation and leaves the specific approach to be justified by the sponsor and discussed with regulators, which is precisely why the early-consultation principle above matters as much as the design principles themselves.
How ICH E17 Relates to the Rest of the ICH Guideline Series
E17 does not stand alone — it sits alongside guidelines that address adjacent parts of trial planning and conduct:
- ICH E5 (Ethnic Factors) supplies the intrinsic/extrinsic factor framework E17 applies proactively during planning rather than retrospectively during bridging.
- ICH E8(R1), General Considerations for Clinical Trials, sets out the broader quality-by-design planning expectations that E17 specializes for the multi-regional case specifically.
- ICH E9, Statistical Principles for Clinical Trials, provides the general statistical framework (randomization, control of bias, pre-specification of analysis) that E17’s regional-consistency considerations sit on top of, rather than replace.
- ICH E6(R2) / ICH E6(R3) govern the conduct-and-oversight standard (GCP) that every site in an MRCT, in every region, must independently satisfy.
- ICH M11, the harmonized clinical trial protocol template, is a practical vehicle for actually documenting an E17-compliant design in a structurally consistent protocol format across regions.
Practical Implications for Research Administrators and Sponsors
For institutions and sponsors participating in or coordinating multi-regional trials, E17 has concrete operational consequences:
- Protocol development needs cross-regional input from the start. Waiting until a protocol is finalized to loop in investigators or regulatory contacts from every planned region risks discovering a region-specific problem with the endpoint, comparator, or eligibility criteria after the design is effectively locked.
- Statistical analysis plans need to state, in advance, how regional consistency will be assessed. Reviewers and IRBs/ethics committees increasingly expect to see this addressed explicitly rather than left implicit.
- Sites participating in an MRCT should expect harmonized, not locally customized, eligibility criteria and endpoint definitions — site-level requests to adapt criteria to local practice need to go through the sponsor’s documented process rather than being handled informally, since ad hoc local variation is exactly what undermines cross-regional consistency.
- Early regulatory consultation should be budgeted for as a real planning step, with lead time for scheduling scientific advice meetings across each participating region’s authority, not treated as optional if timelines are tight.
Frequently Asked Questions
What is a multi-regional clinical trial (MRCT)?
A single clinical trial, run under one protocol, with sites in more than one geographic region, designed from the outset to generate a data package usable to support regulatory submissions in each of those regions rather than requiring separate regional trials.
When was ICH E17 finalized, and where does it apply?
It reached ICH Step 4 (final guideline) in November 2017. It was adopted by Japan’s regulatory authority in June 2018, and the EU and the US FDA in July 2018 (FDA’s Federal Register notice of availability is dated 19 July 2018). As with all ICH guidelines, domestic legal effect depends on each region’s own adoption.
How is ICH E17 different from ICH E5?
ICH E5 (1998) addresses assessing, after the fact, whether clinical data collected in one region can be extrapolated (“bridged”) to support approval in another. ICH E17 applies that same intrinsic/extrinsic-factor thinking proactively, at the trial-design stage, so that a single trial is built from the start to produce data that is interpretable and consistent across all its participating regions.
Does ICH E17 require a specific statistical method for assessing regional consistency?
No. E17 sets out the expectation that a sponsor pre-specify, in its statistical analysis plan, how it will evaluate consistency of treatment effect across regions and how sample size will be allocated across regions — but it does not mandate a single required statistical method. The specific approach is expected to be justified by the sponsor and discussed with regulatory authorities, which is why E17 also emphasizes early regulatory consultation.
Does every clinical trial need to follow ICH E17?
E17 specifically addresses trials designed as multi-regional from the outset. A trial conducted entirely within a single region, or regional trials that are only bridged together after the fact, are governed by other parts of the ICH framework (including ICH E5 for bridging) rather than E17’s MRCT-specific planning principles.
Related CASRAI Resources
See also ICH E6(R2) and ICH E6(R3) for Good Clinical Practice conduct standards, ICH E9 for general statistical principles, ICH E6 vs. ICH E8(R1) for how conduct and design-planning guidelines differ, ICH GCP and ICH (International Council for Harmonisation) for the broader framework, ICH M11: The New Harmonized Clinical Trial Protocol Template, What Is a Clinical Trial?, and the Clinical Research Administration pillar page.
References
- ICH, “E17 General Principles for Planning and Design of Multi-Regional Clinical Trials,” Step 4 final guideline, November 2017 (ich.org)
- US FDA / Federal Register, “E17 General Principles for Planning and Design of Multiregional Clinical Trials; International Council for Harmonisation; Guidance for Industry; Availability,” notice of final guidance, 19 July 2018 (federalregister.gov)
- ICH, “E5(R1) Ethnic Factors in the Acceptability of Foreign Clinical Data,” Step 4, 1998 (ich.org)
- ECA Academy, “FDA publishes final ICH E17 Guidance on multi-regional trials” (gmp-compliance.org)
