Why Seek Parallel Scientific Advice?
A model-based development strategy may need to support decisions in more than one region. The FDA and the European Medicines Agency (EMA) can each provide scientific input, but they operate under different regulatory frameworks and retain independent responsibilities. A sponsor that consults each agency separately may receive feedback at different times, against different versions of the evidence, or in response to differently worded questions. It can then be difficult to tell whether apparent differences reflect distinct scientific views or simply different materials and context.
FDA-EMA parallel scientific advice (PSA) is a coordinated way to seek input from both agencies on selected scientific questions during medicine development. Its value is an opportunity to discuss a common evidence package and learn where agency views converge or differ. The process supports informed planning; it does not create a single regulator, guarantee identical advice, or substitute for either agency’s later review of a formal application.
The precise operational steps, eligibility, document requirements, and interaction format can depend on the agencies’ current procedures and the sponsor’s proposed request. Sponsors should confirm those details with both agencies before relying on a particular process. In particular, “parallel” does not mean that every exchange happens in one meeting or that the agencies issue a jointly negotiated position.
How the Coordination Works
At a practical level, the sponsor identifies a development issue for which advice from both regions would be useful, requests the applicable parallel process, and prepares materials that allow both agencies to consider the same underlying question. The agencies can coordinate aspects of their review and exchange views under the applicable arrangements. They may interact with the sponsor separately; a coordinated or joint discussion may be possible in a particular case, but should not be assumed. The sponsor should confirm the planned format and the rules for sharing submitted information.
The core discipline is to keep the scientific basis aligned. A shared package need not force a single regulatory interpretation. It should, however, make clear which data, assumptions, model version, proposed use, and decision are under discussion. If a regional difference is necessary—for example, because a proposed use or clinical context differs—the sponsor should identify it explicitly rather than allow it to enter unnoticed through inconsistent documents.
Check current FDA and EMA procedures, determine whether the proposed issue is suitable for parallel scientific advice, and confirm the intended agency contacts and interaction format.
Use a consistent description of the development decision, evidence, model assumptions, analyses, and uncertainties. Mark any regional differences and explain why they are relevant.
Submit the questions and supporting materials as directed by the agencies. Track requests, clarifications, and meeting arrangements separately for each authority.
Document each agency’s response, distinguishing agreement, different views, and points that remain unresolved. Convert the record into a transparent development plan.
Coordination depends on the process’s information-sharing arrangements. A sponsor should not assume that material submitted to one agency is automatically available to the other, or that confidentiality permissions are identical across procedures. Confirm the required authorizations, submission channels, and treatment of confidential information. Keep a version-controlled record of what each agency received.
Building a Shared Question and Evidence Package
A coordinated question has the same central scientific issue for both agencies. For instance, a sponsor might ask whether a population pharmacokinetic model, supported by specified clinical data and diagnostics, is an adequate basis for a dose recommendation in a defined population. The package should state the intended decision and model’s proposed role, describe the evidence and limitations, and make clear what advice is being requested. This is not a request for the agencies to approve a model or to guarantee that a future decision will follow the advice.
The sponsor should also check that the question is answerable using the material provided. If the two agencies receive different model versions, different data cutoffs, or different descriptions of the proposed use, divergent feedback may be difficult to interpret. Conversely, identical documents do not require identical agency answers: the agencies may weigh evidence or regulatory context differently.
Worked Example: Keeping a Shared Model Question Aligned
A sponsor is considering whether a population pharmacokinetic model can support a dose recommendation for patients with moderate renal impairment. For this hypothetical request, the sponsor has two model versions, \(M_1\) and \(M_2\), and wants advice from both agencies on one decision.
The sponsor identifies three items that must match in the package: the proposed population, the model version, and the decision under discussion. It selects \(M_2\), which uses the agreed data cutoff, and states the same population and dose-decision context in both agency submissions. The sponsor then asks whether the presented evidence is adequate to support using \(M_2\) for that defined purpose, and which additional evidence or analyses the agencies consider important.
As a version check, suppose the EMA package contains \(M_2\), but a draft FDA package still refers to \(M_1\). The sponsor corrects the draft before submission; it does not describe this as an agency disagreement because no agency has yet expressed a view. If both agencies subsequently comment on \(M_2\) but reach different conclusions, the sponsor can interpret that difference against a common scientific record rather than a package mismatch.
The coordination does not make the advice binding. It makes the question and its evidence easier to compare, which improves the sponsor’s ability to identify what is shared and what requires separate consideration.
Interpreting Agreement and Divergence
Agency responses should be recorded by issue, not compressed into a single overall label such as “the agencies agreed.” Agreement on one point does not imply agreement on every assumption, analysis, or regulatory implication. Likewise, a difference in emphasis is not automatically a direct contradiction. The sponsor should preserve each agency’s wording and distinguish its own interpretation from the agency’s stated advice.
| Record category | What it means | Sponsor’s next step |
|---|---|---|
| Agreement | Both agencies support the same stated point, within the scope of the advice provided. | Record the shared point and its boundaries; do not extend it to unasked questions. |
| Different emphasis | The agencies address related issues but prioritize different evidence, assumptions, or concerns. | Preserve both comments and determine whether the difference changes the evidence plan. |
| Divergent advice | The agencies give materially different recommendations or expectations on the same issue. | Document each position, assess implications for regional plans, and avoid presenting either as a joint conclusion. |
| Unresolved | The available advice does not settle the question, or an agency indicates that more information is needed. | Identify what remains uncertain and what evidence or follow-up could address it. |
Worked Example: Separating Shared Views from Divergent Advice
Suppose, in a hypothetical PSA, both agencies agree that the sponsor should report uncertainty in the renal-impairment prediction. The EMA additionally recommends a sensitivity analysis for a key covariate assumption. The FDA says the current analysis is informative but asks for more evidence before relying on the model for the proposed dose decision.
The sponsor records one area of agreement: both agencies want uncertainty characterized. It records a further EMA request for a sensitivity analysis, and a distinct FDA concern about the evidence needed for the proposed decision. It does not summarize the result as “FDA and EMA agreed that the model is acceptable,” because neither agency gave that shared conclusion.
The sponsor can then plan the sensitivity analysis and assess whether additional evidence is needed to address the FDA concern. If the advice has implications for different regional plans, those implications should be recorded explicitly. The sponsor should preserve the agencies’ separate positions rather than imply that one agency’s advice resolves the other’s concern.
Short Justification: Why a Common Record Helps
The logic of parallel advice is comparative. If two agency responses concern different versions of the question or evidence, a difference between responses has at least two plausible sources: the agency’s assessment or the changed input. Holding the question and supporting record as constant as practicable removes one avoidable source of variation. It does not prove that any remaining difference is caused solely by regulatory philosophy, but it makes the comparison more interpretable.
This is why the sponsor needs both coordination and independent records. Coordination supports a meaningful comparison; separate records preserve which agency said what and the scope of each response. Combining the responses into one paraphrase too early can obscure conditions, qualifications, or genuine differences.
Planning and Common Failure Modes
A PSA takes additional coordination by the sponsor and the agencies. It is most useful when the development decision is important enough to justify a shared evidence package and when the sponsor can act on the resulting advice. It may be less useful if the key question is still undefined, the relevant evidence is changing rapidly, or the sponsor is seeking a binding decision that scientific advice cannot provide.
Worked Example: A Sponsor’s Internal Coordination Plan
A sponsor creates a hypothetical six-week internal work plan before requesting parallel advice. It reserves two weeks to reconcile the scientific package, one week for quality review, and three weeks for internal approval and submission readiness. The arithmetic is \(2 + 1 + 3 = 6\) weeks. These are the sponsor’s planning assumptions, not official FDA or EMA deadlines, and they do not predict how quickly either agency will schedule or complete its process.
During reconciliation, the sponsor finds that one proposed regional summary uses a different data cutoff. It corrects the summary and checks that both versions identify the same model and decision. If the difference cannot be removed because regional contexts genuinely differ, it documents the distinction and its reason rather than concealing it. The plan therefore reduces preventable inconsistencies without assuming that the agencies will give identical advice.
- Assuming “parallel” means joint or harmonized advice. The correction is to treat the agencies as independent authorities and retain separate records of their feedback.
- Sending nominally similar but materially different packages. A different model version or data cutoff can make responses hard to compare. Use version control and identify necessary regional distinctions.
- Calling a partial agreement full consensus. Match each claimed agreement to the specific issue and wording addressed by both agencies.
- Treating advice as approval or a guarantee. Scientific advice informs planning; formal regulatory decisions depend on later submissions, evidence, and applicable requirements.
- Ignoring an unresolved difference. Record the difference and consider whether it affects evidence generation or regional development. Do not silently choose the more convenient response.
A parallel process is one option among separate agency consultations and other available regulatory interactions. Separate consultations may be more practical when the issues are region-specific or the sponsor does not need coordinated consideration. Parallel advice offers a structured opportunity to compare views on shared questions, but involves coordination and does not remove regional requirements. The choice depends on the question, evidence, and intended use of the advice.
Check Your Understanding
Answer each question, then compare your reasoning with the solutions.
- Does parallel scientific advice mean the FDA and EMA issue one binding joint decision?
- Why should a sponsor check that the agencies receive consistent model versions and data cutoffs?
- If both agencies agree on an uncertainty analysis but differ on whether the model supports a dose decision, how should the sponsor record the outcome?
- A sponsor reserves one week for document review and two weeks for internal approvals. What is the total internal time, and should it be described as an agency deadline?
- When might separate consultations be more suitable than a parallel process?
Solutions. 1. No. The agencies retain independent authority, and scientific advice is not a binding joint decision. 2. Consistency makes the responses more comparable and reduces avoidable ambiguity about whether a difference reflects the evidence package or the agencies’ assessments. 3. Record agreement only on the uncertainty analysis and document the dose-decision positions separately, including any unresolved issue. 4. The sponsor’s internal estimate is \(1 + 2 = 3\) weeks. It is not an FDA or EMA deadline. 5. Separate consultations may suit region-specific questions or situations where coordinated consideration of a shared issue is not needed.