In practice, the pharmaceutical company may still provide the drug, funding, operational support, safety information, and regulatory expertise — while being explicitly forbidden from becoming the trial’s real sponsor.
This is where the mythology becomes expensive.
The investigator or institution remains the sponsor-investigator, with responsibility for protocol design, data management, ethics approvals, safety reporting, and overall trial conduct. Medical Affairs must support the research without quietly converting scientific collaboration into outsourced commercial strategy. The distinction is not cosmetic. It determines who carries legal responsibility, who controls the data, who owns the publication decision, and whether the study can survive regulatory scrutiny once the friendly launch meeting is over.
I have seen enough medical affairs programmes to know that the most dangerous IIT failures rarely begin with an obvious attempt to manipulate science. They begin with blurred roles, optimistic assumptions, and a contract that reads like procurement paperwork rather than a map of accountability.
An IIT is not a company-sponsored study with a more flattering label. The investigator owns the trial; Medical Affairs must respect that ownership in operational detail, not just in slide-deck language.
The sponsor-investigator boundary is not a branding exercise
The central question in any investigator-initiated trial is deceptively simple: who is the sponsor?
Under the ICH E6 framework and applicable national regulations, the independent investigator or institution serving as sponsor assumes full legal, ethical, operational, and regulatory responsibility for the research. That includes designing the protocol, obtaining institutional review board or ethics committee approval, maintaining the trial documentation, managing data, reporting safety information, and ensuring that the study is conducted according to Good Clinical Practice.
The pharmaceutical company may fund the trial or provide investigational product. It may offer technical information about the compound, review a draft protocol for feasibility or safety concerns, and support agreed activities under a formal contract. None of that automatically transfers sponsorship.
This is the first place where corporate language tends to do damage. A company may describe itself as a “strategic partner,” a “research sponsor,” or a “programme owner” in internal communications. Those phrases may sound harmless in a town hall. They become less harmless when they leak into contracts, operating procedures, investigator correspondence, or regulatory records. If the company behaves as though it controls the trial, it may create evidence that contradicts the formal sponsor-investigator structure.
A credible Medical Affairs operating model therefore draws a hard line between support and control.
What the investigator or institution owns
The sponsor-investigator is accountable for the scientific and operational architecture of the study. That normally includes:
- framing the research question and study objectives;
- developing and approving the protocol;
- securing ethics committee or IRB approvals;
- obtaining informed consent in accordance with applicable requirements;
- maintaining the Trial Master File;
- ensuring appropriate data collection, management, and analysis;
- overseeing investigators and participating sites;
- reporting safety information within required timelines;
- registering the trial on an appropriate public registry before execution;
- preparing and submitting required regulatory documentation.
The exact obligations vary by jurisdiction, product, and study design. The underlying principle does not: the entity that serves as sponsor cannot delegate accountability merely because another organisation supplies money, product, or expertise.
The company’s role must be defined with equal care. Medical Affairs can review whether a proposed study raises product safety concerns. It can provide the investigator with current scientific and regulatory information. It can assess whether the requested support is feasible, lawful, and aligned with the company’s research strategy. It can insist that agreed pharmacovigilance processes are followed.
It cannot approve the protocol as though it owns the science, dictate the study outcome, suppress an unfavourable interpretation, or steer publication decisions through commercial channels.
That last point matters because the temptation is rarely dramatic. No one needs to send an email saying, “Please produce a positive result.” Influence usually arrives in softer packaging: a request to adjust the endpoint, narrow the patient population, delay a manuscript, add a preferred analysis, or remove an inconvenient comparator. Each request can be defended as reasonable in isolation. Together, they can turn investigator-led research into a carefully managed echo chamber.
Regulatory thresholds: an IIT is not automatically exempt
The phrase “investigator-initiated” describes who originates and sponsors a study. It does not provide a universal regulatory exemption.
In the United States, drug studies may qualify for exemption from the Investigational New Drug requirements under 21 CFR § 312.2(b) only if they meet five specific criteria. Those criteria include, among other conditions, that the investigation is not intended to support a new indication or a significant change to product labelling or advertising, and that it does not involve increased risk to patients.
That distinction should be embedded in the initial feasibility assessment, not discovered after funding has been approved.
A study can be academically independent and still require an IND. It can receive no commercial funding and still trigger regulatory obligations. It can use an approved product and still involve a design, population, dose, route of administration, or risk profile that changes the regulatory analysis. The label “non-commercial” does not make the regulatory question disappear; it merely makes people more likely to postpone asking it.
Medical Affairs oversight should therefore begin with a structured regulatory classification of the proposal. The review should consider:
- the product and its approved status in each relevant jurisdiction;
- the proposed indication, population, dose, route, and duration;
- whether the study is intended to support a new indication or labelling change;
- whether the intervention introduces increased risk;
- whether the study involves a device or device component;
- the sponsor-investigator’s ability to meet reporting and documentation duties;
- the requirements of local ethics committees and competent authorities;
- the relationship between the protocol and any existing company commitments.
For device-related research, 21 CFR Part 812 may become relevant. Human subject protection requirements under 21 CFR Parts 50 and 56, financial disclosure under Part 54, and applicable requirements under 45 CFR Part 46 may also enter the assessment. In Europe, Regulation (EU) No 536/2014 provides the framework for clinical trials, with national implementation and operational requirements shaping how the study is submitted and conducted.
The point is not to turn every Medical Affairs colleague into a regulatory lawyer. It is to stop the organisation from treating legal classification as an administrative footnote. An IIT strategy that starts with scientific enthusiasm and adds compliance at the end is not a strategy. It is a delayed escalation.
The global layer makes shortcuts even less attractive
Cross-border investigator-led research creates additional complexity because sponsorship, ethics review, safety reporting, data protection, importation, and public registration may operate through different systems. A protocol that appears manageable in one country can become operationally fragile once multiple jurisdictions, institutions, and reporting calendars are involved.
The sponsor-investigator and supporting company should agree early on:
1. Which party is responsible for each regulatory submission.
2. Which safety events must be communicated to the company, and within what timeframe.
3. How investigational product will be supplied, stored, reconciled, and returned.
4. Which organisation maintains the authoritative trial records.
5. How protocol deviations, serious breaches, and urgent safety measures are handled.
6. Which registry will be used and who is responsible for keeping the record current.
7. How local requirements interact with the global study plan.
The updated ICH E6(R3) Good Clinical Practice guidelines move the industry further toward quality-by-design and risk-based oversight. That does not reduce accountability. It makes poor prioritisation more visible. A risk-based approach is not permission to document less, monitor less, or ask fewer questions. It is a requirement to identify what can compromise participant protection or the reliability of results — and to put controls around those risks rather than applying ritualistic process to everything equally.
Pharmaceutical support needs a real contract, not a ceremonial signature
Funding and drug supply are often described as support, but support still creates obligations. A company that contributes resources to an IIT needs a formal written agreement that describes those obligations with enough precision to withstand a disagreement.
At minimum, the agreement should address:
- the respective roles and responsibilities of the sponsor-investigator, institution, and company;
- the scope and conditions of financial support;
- provision, storage, accountability, and disposal of investigational product;
- compliance with anti-bribery and anti-kickback laws;
- data ownership, access, retention, and security;
- publication and presentation rights;
- safety reporting responsibilities and timelines;
- audit and inspection cooperation;
- handling of protocol amendments and deviations;
- termination conditions;
- milestone-based disbursement of funds.
This is not bureaucratic decoration. It is the operating system of the collaboration.
Milestone-based payments can protect both parties when they are linked to objective deliverables such as ethics approval, site activation, completion of agreed recruitment milestones, database lock, or submission of a final report. They become problematic when payment is tied to a favourable outcome, a preferred interpretation, or a publication decision controlled by the company. The difference is obvious when written down, which is precisely why it should be written down.
Data ownership requires similar discipline. The company may need access to safety data or agreed study outputs. It may need to review information for regulatory reporting. It may have legitimate interests in understanding product performance. None of those interests should become a blank cheque for controlling the entire dataset.
A sound agreement answers practical questions that are often postponed until the relationship becomes uncomfortable:
- Can the investigator publish negative or inconclusive findings?
- How long may the company review a manuscript?
- What can be removed for confidential commercial information?
- Who decides whether a safety signal requires notification?
- Can the investigator access the complete dataset?
- What happens if the company terminates support?
- Who retains records if the institution changes personnel?
- How are disputes over authorship or analysis resolved?
If the agreement cannot answer those questions, the collaboration is not transparent. It is merely hopeful.
The contract should preserve scientific independence before anyone needs to defend it. Once the conflict appears, goodwill is a remarkably weak control.
Medical Information and Pharmacovigilance must not become informal governance
Companies supporting IITs often provide medical information services, safety data, and product expertise. That support can be valuable, especially when investigator teams have limited experience with the product’s risk profile or global safety environment. But informal advice must not evolve into unrecorded decision-making.
Medical Information should provide accurate, balanced, scientifically supported responses within its defined remit. Pharmacovigilance teams should receive safety information according to agreed timelines and procedures. Medical Affairs should ensure that these interfaces are documented and understood by the investigator team.
The practical danger is the side conversation: a call in which an investigator raises a possible adverse event, a message suggesting that a protocol deviation is “probably not significant,” or a request for an informal view on whether a report is necessary. None of these exchanges becomes harmless simply because they occur outside the formal study platform.
A defensible IIT framework treats communication channels as part of the control environment. If a piece of information could affect participant safety, data integrity, reporting, or trial conduct, it belongs in the agreed process.
Applying ICH E6(R3): quality by design without quality theatre
The adoption of ICH E6(R3) on January 6, 2025, with implementation activity expected in the European Union in July 2025 and FDA implementation guidance in September 2025, reinforces a direction the more serious parts of the industry have been moving toward for years: quality should be designed into the study rather than inspected into existence at the end.
For IITs, that principle has special value. Investigator-led research may involve smaller teams, uneven infrastructure, academic timelines, and variable experience with sponsor responsibilities. A quality-by-design approach helps identify the features of the protocol that genuinely affect participant protection and the credibility of the evidence.
The first step is to identify critical-to-quality factors. Depending on the study, these may include:
- correct eligibility assessment;
- accurate exposure and dosing records;
- timely capture and escalation of safety events;
- reliable primary endpoint measurement;
- informed consent before study procedures;
- protection of the randomisation or blinding process;
- complete and traceable source data;
- appropriate handling of protocol deviations;
- statistical analysis that matches the prespecified plan.
Once those factors are identified, the oversight model should focus on the risks that could compromise them. That may mean targeted monitoring, central review of selected data, focused training, or predefined escalation triggers. It does not mean abandoning the Trial Master File or treating documentation as optional.
A risk-based model also requires proportionality. An early-phase, multi-centre interventional study with a vulnerable population will not need the same controls as a low-risk observational project using existing records. Applying identical monitoring procedures to both may create the comforting appearance of consistency while wasting resources on low-value activity.
A practical Medical Affairs oversight model
Medical Affairs can support quality without taking over the trial by separating four activities that are too often blended together.
1. Scientific assessment
The review should test whether the research question is meaningful, answerable, and relevant to patient care. A fashionable topic is not necessarily a useful one. The proposal should have a credible rationale, an appropriate design, feasible recruitment assumptions, and endpoints capable of generating interpretable evidence.
This is the point at which scientific merit review must remain separate from commercial preference. A study should not become “high priority” merely because its likely result is convenient.
2. Compliance and feasibility assessment
The team should examine whether the investigator or institution can meet sponsor obligations. A compelling scientific concept is not enough if the proposed sponsor cannot manage safety reporting, maintain essential records, obtain approvals, or supervise participating sites.
This is also where the company should assess whether the requested support is lawful, proportionate, and consistent with internal policies. Commercial personnel should not participate in the review or approval process for IIT grants. That separation is not an insult to the commercial organisation; it is a safeguard against conflicts that everyone will later pretend were unforeseeable.
3. Operational oversight
Once support is approved, Medical Affairs should monitor agreed milestones and risk indicators without directing day-to-day trial conduct. Oversight may include scheduled status updates, review of safety communication processes, confirmation of registry and ethics obligations, and escalation of material deviations.
The question is not whether the company can see everything. It is whether the company can identify when the study is no longer operating within the agreed scientific, ethical, or contractual boundaries.
4. Close-out and evidence use
At the end of the study, the company should receive the agreed reports and safety information. It may assess how the findings contribute to medical understanding, future research, or responsible scientific communication. It should not rewrite the study’s conclusions to fit a pre-existing narrative.
Negative results are still evidence. Inconclusive results are still information. A trial that fails to support the preferred hypothesis has not necessarily failed as a research investment. It may have exposed an ineffective assumption before that assumption reached clinical practice.
Scientific merit review is where commercial influence usually hides
The most sensitive stage of managing investigator-initiated research is often not publication. It is selection.
Companies receive more proposals than they can support. They need a process for deciding which studies merit funding, drug supply, or other assistance. That process should be based on scientific merit, patient relevance, feasibility, ethical acceptability, and the investigator’s ability to execute the work.
It should not become a disguised market access or promotional planning meeting.
A robust IIT scientific merit review can ask:
- Does the question address an important evidence gap?
- Is the proposed design capable of answering it?
- Are the endpoints clinically meaningful?
- Is the population appropriately defined?
- Does the investigator have relevant expertise and infrastructure?
- Are recruitment and follow-up realistic?
- Are the statistical assumptions credible?
- Does the study duplicate existing evidence without a clear rationale?
- Are participant risks proportionate to the expected knowledge gain?
- Can the sponsor-investigator meet the regulatory and operational obligations?
- Are publication and data access arrangements acceptable?
- Are there conflicts of interest that require mitigation?
The review should be documented, consistent, and independent of sales targets. Decisions should be explainable without resorting to phrases such as “strategic fit” that mean everything and therefore protect nothing.
That does not mean Medical Affairs must pretend that strategy is irrelevant. Medical Affairs has legitimate priorities: unmet medical needs, evidence gaps, safety questions, clinical practice variation, and the responsible generation of real-world evidence. The problem begins when those priorities are translated into pressure for a predetermined result.
The cleanest internal test is simple: would the company still consider supporting the proposal if it produced a neutral or unfavourable finding? If the answer is no, the project may be a commercial activity wearing a scientific costume.
Building an IIT strategy that survives scrutiny
An effective medical affairs IIT strategy does not begin with the number of proposals funded. That is an easy metric and a poor one. It begins with the quality of the research questions, the integrity of the selection process, and the reliability of the evidence produced.
The operating model should define:
- who may submit or sponsor a proposal;
- how unsolicited concepts are handled;
- who performs scientific and compliance review;
- how commercial functions are excluded from approval decisions;
- how conflicts of interest are identified;
- which support types are available;
- how budgets and milestones are assessed;
- how safety and medical information interfaces operate;
- how protocol amendments are reviewed;
- how results and publications are tracked;
- how lessons from completed studies influence future oversight.
The organisation should also distinguish investigator-initiated trials from other evidence-generation activities. An IIT is not simply another route to a Phase IV study. In a company-sponsored Phase IV programme, the company remains responsible for the study as sponsor. In an IIT, the investigator or institution carries that responsibility. Confusing these models creates unrealistic expectations about control, timelines, data access, and accountability.
The distinction matters operationally. A company cannot demand the same degree of direction over an IIT that it would exercise in a company-sponsored trial. It can, however, decide whether to provide support, define the conditions of that support, and withdraw from a collaboration when those conditions are not met — provided the decision does not amount to retaliation for independent scientific conclusions.
That is the balance. Independence does not mean the investigator operates without standards. Compliance does not mean the company becomes the hidden sponsor.
The uncomfortable conclusion
Investigator-initiated trials are valuable precisely because they can ask questions that commercial programmes may overlook. They can examine real clinical uncertainty, populations excluded from pivotal studies, treatment patterns, safety concerns, and outcomes that matter to clinicians rather than quarterly presentations.
But independence is not created by removing the company’s name from the sponsor field. It is created through governance: clear accountability, transparent contracts, defensible scientific review, disciplined safety processes, reliable documentation, and publication rights that survive an inconvenient result.
Medical Affairs should be the function that protects that architecture. Not by standing at a distance and applauding “collaboration,” but by defining where support ends and control begins.
The actionable takeaway is less glamorous than the usual strategy language: write the boundary down, resource the investigator properly, document every material decision, and reject any review process that cannot tolerate a negative answer. If the study only works when the result is commercially useful, it was never an independent trial in the first place.