Phase I transparency under the EU Clinical Trials Regulation: what CTIS publishes, when, and how sponsors defer or redact

Phase I studies sit at the point where the Clinical Trials Regulation's transparency ambitions collide most directly with sponsors' commercial sensitivities. First-in-human data, dose-finding rationales and quality information on unlicensed molecules are exactly the material a sponsor most wants to protect, and exactly the material the public disclosure regime is designed to surface. This overview sets out which documents and data become public for a Phase I trial in the Clinical Trials Information System (CTIS), the timing rules, and the two levers sponsors have to manage disclosure: deferral and redaction. It closes with concrete examples of how sponsors have actually applied those levers in trials already visible in CTIS.

The legal baseline: Article 81 of Regulation (EU) 536/2014

The starting principle is disclosure. Regulation (EU) 536/2014 requires the EMA, with the Member States and Commission, to maintain the EU database (the public-facing layer of which is CTIS), and Article 81(4) states that the database "shall be publicly accessible unless, for all or part of the data and information contained therein, confidentiality is justified on any of the following grounds" ⁷⁸:

protecting personal data (Article 81(4)(a)) ⁷⁸;
protecting commercially confidential information, "in particular through taking into account the status of the marketing authorisation for the medicinal product, unless there is an overriding public interest in disclosure" (Article 81(4)(b)) ⁷⁸;
protecting confidential communication between Member States in relation to the preparation of the assessment report (Article 81(4)(c)) ⁷⁸;
ensuring effective supervision of the conduct of a clinical trial by Member States (Article 81(4)(d)) ⁷⁸.

Two further paragraphs frame the timing and the hard limits. Article 81(5) provides that, absent an overriding public interest, "data contained in the application dossier shall not be publicly accessible before the decision on the clinical trial has been made" ⁷⁸. Article 81(7) is absolute: "No personal data of subjects shall be publicly accessible" ⁷⁸.

The recitals set the tone a reviewer should keep in mind. Recital 67 says the database should be publicly accessible and searchable, with related documents (the summary, the layperson summary, the protocol and the clinical study report) linked by the EU trial number, and that information should be public "unless specific reasons require that a piece of information should not be published" to protect private life and personal data ⁷⁸. Recital 68 is the key constraint on the commercial-confidentiality argument: "in general the data included in a clinical study report should not be considered commercially confidential once a marketing authorisation has been granted" ⁷⁸. In other words, the CCI shield weakens sharply once the product is authorised.

Which documents and data become public

For a Phase I trial, the material that flows to the CTIS public domain falls into three buckets:

Structured data. The trial's main characteristics (EU trial number, title, sponsor, condition, phase, population, Member States concerned, status and dates) are structured fields published under the transparency timelines. Conclusions and decision outcomes of an application, together with the corresponding dates, are made publicly available at the time of decision for every trial ⁷².

Application-dossier documents. These include the protocol and protocol-related documents, the Investigator's Brochure, the Investigational Medicinal Product Dossier (both the quality part, IMPD-Q, and the safety/efficacy part), GMP documents and labels in Part I, and Part II documents such as the recruitment arrangements and the subject information sheet / informed consent form ⁷¹. Not everything the sponsor uploads is published: notably, the assessment report on Part I and Part II is not published, as detailed in the Annex to the Revised CTIS transparency rules; only the conclusions, decision outcomes and dates are disclosed at the time of decision ⁷².

Results. The summary of results and the layperson (lay) summary are published, as is the clinical study report where applicable. Under Article 28, trial participants are told at consent that "the summary of the results of the clinical trial and a summary presented in terms understandable to a layperson will be made available in the EU database ... irrespective of the outcome of the clinical trial, and, to the extent possible, when the summaries become available" ⁷⁰.

The publication rules themselves are not spelled out in the Regulation. They live in the Revised CTIS transparency rules and, in operational detail, in the Guidance document on how to approach the protection of personal data and commercially confidential information while using CTIS and its Annex I, which the CTR Q&A identifies as the governing texts for CTIS disclosure ⁷². The revised rules "foresee the disclosure of structured data and key documents of public interest as per timelines based on the trial category, trial phase and population age" ⁷².

Timing: when publication happens

Three timing anchors matter for a Phase I trial.

At the decision. Application-dossier documents in scope of publication are published after the decision on the application, consistent with Article 81(5)'s bar on pre-decision disclosure ⁷⁸. The EMA guidance encourages sponsors to submit a lightly redacted "for publication" version so that documents can be published promptly after the decision rather than held back ⁷².

Results after end of trial. A summary of results is due within one year of the end of the trial for an adult trial, and within six months for a paediatric trial; where a paediatric trial falls outside Article 46(1) of the Paediatric Regulation and it is not scientifically possible to meet the six-month deadline, the summary is due no later than 12 months after trial end ⁶⁸. The layperson summary accompanies the summary of results and follows the same submission timing ⁶⁸.

Intermediate analyses. Where the protocol provides for an intermediate data analysis before the end of the trial, a summary is due within one year of the intermediate analysis date, with exceptions that are highly relevant to early-phase work: where the blind must be maintained (for example, an independent DSMB analysis); where the protocol sets clear criteria for how to continue the trial (the guidance's own example is a dosing-regimen decision in an early-phase trial); and where there are justified reasons why the one-year summary is not possible ⁷².

The two levers: deferral and redaction

Sponsors manage disclosure through two distinct mechanisms.

Deferral (the trial-category mechanism)

At the time of the initial application, the sponsor can set deferral settings to delay publication of clinical trial data and documents, and thereby protect commercially confidential information without having to redact the documents themselves ⁷¹. The permitted deferral is driven by the trial category the sponsor selects (subject to review by the Member States concerned), based in particular on the marketing-authorisation status of the investigational product ⁷¹:

Category 1: phase 0 and phase I trials; bioequivalence and bioavailability trials; similarity trials for biosimilars; and equivalence trials for combination or topical products where PK/PD studies are not possible ⁷¹.
Category 2: phase II and phase III trials ⁷¹.
Category 3: phase IV and low-intervention trials ⁷¹.

Phase I therefore sits in Category 1, the category associated with the greatest commercial sensitivity and, correspondingly, the longest permitted deferral. Where a protocol with a multiphase or adaptive design spans both Category 1 and Category 2, it is treated as Category 2 ⁷¹.

The category determines the length of deferral available. The precise maximum deferral periods per category, and the events from which they run, are set out in the Revised CTIS transparency rules and its Annex I rather than in the Regulation or the sponsor-facing guides reviewed here; a reviewer confirming a specific deferral clock for a Phase I trial should read those figures directly from Annex I. What the guidance does make explicit is the mechanics around deferral:

If a deferred trial is authorised, the data and documents subject to deferral are not published at the time of the decision ⁷¹.
If a deferred trial is rejected by all Member States concerned receiving the application, the information is still published on the same deferral clock as for an authorised trial, with the date of the last Member State's decision on a full Part I and Part II application treated as the end of trial for that purpose ⁷¹.
Member States concerned may delay publication of their own documents (Part I and Part II assessment reports and requests for information) in line with the sponsor's deferral timelines ⁷¹.

There is an important carve-out that directly limits Category 1 deferral: for Category 1 trials in paediatric populations and/or forming part of a paediatric investigation plan, it is not possible to defer publication of the main characteristics, notifications, the summary of results and intermediate data analyses ⁷¹. Trials in a declared public health emergency of international concern cannot be deferred at all; their protocol is published at the time of the decision on the application ⁷²⁷¹.

Redaction (the "for publication" / "not for publication" mechanism)

The alternative and increasingly preferred lever is redaction. CTIS lets the sponsor upload two versions of a document: a redacted version "for publication" and an unredacted version "not for publication," the latter carrying everything the assessors need ⁷¹. The EMA's stated policy preference is to minimise deferral in favour of light redaction: sponsors are "encouraged to submit trial documents 'for publication' with a minimum amount of redactions limited to personal data and commercially sensitive information," so documents can be published soon after the decision, while the redacted version must "remain meaningful to the public, including potential trial participants and health care professionals" ⁷². The guidance is explicit that this is aimed squarely at early development: redaction "would enable the earliest publication of trial documents and at the same time keep sponsors confidence in using EU for clinical trials, especially for early development where sensitivities are highest," and warns that "extensive deferrals could significantly reduce the utility of clinical data in CTIS" ⁷².

For personal data the rule is stricter than a preference. Personal data in uploaded documents must be kept to the minimum; where it is needed (for example pseudonymised subject identifiers) it goes in the "not for publication" version and is anonymised in the "for publication" version ⁷¹. This operationalises the absolute Article 81(7) bar on public access to subjects' personal data ⁷⁸.

The shift from deferral to redaction under the revised rules

The regime is mid-transition. Following publication of the Revised CTIS transparency rules, and pending their full technical implementation in CTIS, sponsors filing an initial application may already follow the revised approach: rather than setting deferrals, they provide "for publication" and "not for publication" versions only for the documents in scope of the revised rules (listed in Annex I), and for document types that will no longer be published they upload a placeholder page using the wording suggested in the ACT EU Q&A Annex I ⁷¹. Document types dropped from publication under the revised rules include final assessment reports and decision letters ⁷¹. For "historical" trials already in CTIS before technical implementation, structured data fields for all categories will be published under the revised timelines, but documents contained in applications submitted before implementation will not be published, regardless of the prior deferral or publication status ⁷¹.

The net effect for a Phase I sponsor: deferral by trial category remains legally available, but the direction of travel, and the EMA's express encouragement, is toward publishing promptly after the decision with targeted redactions for CCI and personal data rather than blanket delay.

How sponsors have actually applied this in CTIS

Trials already visible in CTIS show both mechanisms in use, and the redaction language clusters heavily in early-phase oncology and first-in-human studies where sensitivities are highest.

Partner Therapeutics / Merus, in a Phase I/II first-in-human solid-tumour protocol, reserves the right to delay a proposed publication to protect intellectual property and to "require that such information or data are removed from the proposed publication" ⁹¹.
Novartis, in a Phase I/II first-in-human oncology protocol, states that any confidential information in a proposed publication "will be removed at Sponsor's request," and that where patentable subject matter is disclosed, publication "shall be delayed" for up to 90 days ⁹³.
AbbVie filed a protocol on CTIS explicitly as a "Redacted" public version marked "Confidential Information," the visible artefact of the two-version approach ¹⁰⁰.
Merck Sharp & Dohme attached an instruction sheet to a recruitment document stating that where redaction is needed, "two versions must be created: one redacted version (for publication) and one non-redacted version (not for publication)" ¹⁰⁴, a near-verbatim application of the CTIS mechanism.
Takeda uses the same construction in a Phase III protocol: up to a 90-day publication delay for patentable subject matter, with confidential information "removed at Takeda's request" ⁹⁹.
Mundipharma, in a Phase II protocol, allows publication to be "delayed for a further 6 months or until its intellectual property rights have been secured," and may require proprietary information "to be removed prior to such publication" ¹⁰¹.
Pfizer posts "Public Disclosure Synopses" from which "any data that could be used to identify individual patients has been removed," and reserves the right to remove previously undisclosed confidential information before disclosure ⁸⁹.
Janssen Cilag International controls timing rather than content, providing that disclosure of final results occurs only after study end, within 18 months of the study end date ⁹².

The pattern is consistent with the guidance: sponsors are publishing redacted public versions rather than relying solely on category-based deferral, and the strongest confidentiality language appears in the Phase I / first-in-human protocols, exactly where the Regulation places the trial in Category 1 and where CCI sensitivities peak before any marketing authorisation exists to erode them under Recital 68 ⁷⁸.

Practical takeaways for a Phase I sponsor

Phase I is Category 1, which affords the longest deferral but also draws the most scrutiny; if any part of the programme is paediatric or PIP-linked, deferral of main characteristics, notifications, results and intermediate analyses is off the table ⁷¹.
Deferral protects CCI without touching the documents but delays public value; redaction publishes sooner and is the EMA's stated preference for early development, provided redactions stay limited to personal data and genuine CCI and the public version remains meaningful ⁷².
Personal data handling is not discretionary: minimise it, keep it in the "not for publication" version, anonymise the public version, and remember Article 81(7)'s absolute bar ⁷⁸⁷¹.
The CCI argument has a shelf life. Once a marketing authorisation is granted, clinical study report data are generally not commercially confidential, so a deferral strategy built around a molecule that is progressing toward authorisation should be planned with that erosion in mind ⁷⁸.
The exact deferral clocks by category are defined in the Revised CTIS transparency rules and its Annex I; confirm the specific maximum period and its start event there before relying on a deferral in a submission strategy ⁷².

Phase I Transparency Under the EU Clinical Trials Regulation and CTIS