Vitalis Therapeutics Inc. — The four formal FDA meetings across the VitaFlow (VTX-401) program: meeting types and their procedural timelines, how a meeting actually works including preliminary responses, what each meeting changed, and what agreement does and does not mean.
1. Four Meetings in Six Years
| Meeting | Type | Held | Preceded |
|---|---|---|---|
| Pre-IND | Type B | 14 May 2023 | Gate 1 |
| End-of-Phase 1 | Type B | 22 Apr 2024 | Gate 3 |
| End-of-Phase 2 | Type B (EOP) | 19 May 2026 | Gate 4 |
| Pre-NDA | Type B | 10 Aug 2028 | Gate 5 |
A gate is a decision taken on evidence. An agency meeting is where a specific, expensive class of uncertainty gets resolved — what the regulator will accept. Holding the meeting after the gate would mean authorizing money against an assumption about the agency's position rather than a record of it.
The clearest case is End-of-Phase 2 in May, Gate 4 in June. Six weeks apart, and the wrong way round would have committed $125,800,000 to a pivotal design the agency had not yet commented on.
Four formal meetings across 8 years is not many, and that is deliberate. Meetings are expensive in preparation time and in the schedule float they consume — a Type B meeting is sixty days from request to date, and the briefing package is due thirty days before that. Ninety days of a program's calendar for one conversation.
2. Meeting Types
| Type | Request to meeting | When it is used | Note |
|---|---|---|---|
| Type A | 30 days | Program is stalled or on a critical path | Dispute resolution, clinical hold response, post-CRL. The meeting you never want to need. |
| Type B | 60 days | Development milestones | Pre-IND, End-of-Phase 1, Pre-NDA. The scheduled architecture of a program. |
| Type B (EOP) | 70 days | End-of-Phase 2 specifically | Longer because the scope is wider — pivotal design, endpoints, statistics, safety database size, CMC expectations. |
| Type C | 75 days | Anything else | Broad scope, lowest urgency. Package due 47 days before the meeting. |
| Type D | 50 days | A narrow set of issues (PDUFA VII) | No more than two focused topics, three disciplines, five questions. Package due with the request. |
| INTERACT | 75 days | Earliest stage, novel products | Precedes pre-IND. For programs facing questions with no existing guidance. |
The type is not a priority setting the sponsor chooses. It is a description of what the meeting is for, and the timelines follow from that description.
All four of this program's meetings were Type B, which is what a program running to plan looks like. A Type A meeting anywhere in this log would indicate something had gone badly wrong — a clinical hold, a dispute, a Complete Response Letter to respond to. The absence of one is itself a finding.
3. How a Meeting Actually Works
| Step | Who | Detail |
|---|---|---|
| Request with questions | Sponsor | The request states the meeting type, the objectives and the specific questions. A request without numbered questions produces a discussion. |
| Grant or deny | FDA | The agency confirms type, format and date — and may convert the type if the scope does not match what was requested. |
| Briefing package | Sponsor | 30 days before a Type A or B meeting. Background, the questions, the sponsor's position on each, and the data supporting it. The package is the meeting. |
| Preliminary responses | FDA | The agency answers the questions in writing, before the meeting. |
| Do we still need to meet? | Sponsor | Within 3 calendar days of receiving preliminary responses, the sponsor says whether the meeting is still required. |
| Meeting | Both | Face-to-face, virtual face-to-face, teleconference, or written response only. Presentations are usually unnecessary — the content is already in the package. |
| Official minutes | FDA | Issued within 30 days. The agency's minutes are the record, not the sponsor's. |
FDA answers the questions in writing before the meeting happens. Preliminary responses arrive ahead of the scheduled date, and the sponsor then has three calendar days to say whether the meeting is still needed at all.
That inverts the usual expectation of what a meeting is. The written answers are the substance; the meeting exists to resolve whatever the written answers left ambiguous. A sponsor that reads the preliminary responses, finds every question answered, and cancels has not wasted the process — it has used it correctly.
The consequence for how a package is written is direct. Since the agency answers in writing, from the package, without the sponsor in the room, the package has to stand alone. A question that depends on being explained in person will be answered as written, not as intended. That is why a meeting request carries numbered questions with the sponsor's position stated on each: the agency is responding to the position, and an unstated position gets no response.
Programs get into trouble years later by acting on what they remember agreeing rather than on what the minutes say was agreed.
4. What Each Meeting Changed
The only measure of a meeting that matters.
Pre-IND (Type B) — 2024-08-14
Scope: Nonclinical package sufficiency; first-in-human design; CMC expectations
End-of-Phase 1 (Type B) — 2025-07-22
Scope: Phase 2 dose selection and dose-ranging design
End-of-Phase 2 (Type B EOP) — 2026-05-19
Scope: Pivotal design, co-primary endpoints, CV safety expectations, statistical plan
Pre-NDA (Type B) — 2028-08-10
Scope: NDA content and format, integrated summaries, labeling strategy
The Pre-IND added a toxicology study, eleven weeks and roughly $1.2M. End-of-Phase 1 added a dose arm and 120 participants. End-of-Phase 2 added a cardiovascular sub-study worth $7,200,000.
None of those were losses. Each was a requirement the program would have met eventually — at Gate 5, at filing, or in a Complete Response Letter — discovered at the cheapest possible moment instead. A meeting that changes nothing was either unnecessary or was not asked the right questions.
5. What Agreement Does and Does Not Mean
| Is a meeting outcome binding on FDA? | No. Advice reflects the agency's position on the information available at the time. |
| Can the agency change position later? | Yes — if new data emerges, if the program changes, or if the science moves. |
| So what is it worth? | A great deal. Departing from prior written advice requires the agency to explain why, and it puts the sponsor's design on the record as having been reviewed. |
| What if the sponsor departs from the advice? | Permitted, and it must be justified in the submission. An unexplained departure from agreed design is a review question waiting to be asked. |
This program departed from agreed advice once — the dose-escalation schedule extended from 16 to 20 weeks after the End-of-Phase 2 discussion, recorded in TPP v3.0 as the only conceded target. It was documented at the time rather than discovered at filing.
6. What the Program Manager Contributes
| Contribution | In this program |
|---|---|
| Sequencing meetings against gates | End-of-Phase 2 in May 2026, Gate 4 in June 2026. Six weeks, and the right way round. |
| Protecting the ninety days | Sixty from request to meeting, thirty for the package before it. A meeting requested late is a gate delayed. |
| Insisting on numbered questions | The agency answers questions. A request describing a topic gets a discussion. |
| Budgeting for the answer, not the meeting | CR-02 was not the cost of the End-of-Phase 2 meeting. It was the cost of what the meeting revealed — $7,200,000, drawn from contingency because the reserve exists for exactly this. |
| Reconciling minutes against internal notes | Within the 30-day window, while anyone still remembers the conversation. |
Going into End-of-Phase 2, the program held contingency against precisely this. The $7,200,000 that came out of it was drawn the same week — not because the outcome was predicted, but because the category of outcome was, and reserve had been sized for it.
7. The Meetings Not Requested
Four meetings in 8 years is a deliberate number. The log above records what was asked; this section records what was considered and not asked, because in agency interaction the decisions not to engage carry as much weight as the engagements.
| Considered | When | Not requested because |
|---|---|---|
| Type C on the statistical analysis plan | Stage 4, before database lock | The estimand strategy and the testing hierarchy had already been agreed in the End-of-Phase 2 minutes. Reopening a settled question invites a fresh position on it, and the sponsor had nothing to gain from a second answer to a question already answered favorably. |
| Type B on a comparative tolerability claim | Stage 4 | The claim depended on data sitting fourth in a six-position testing hierarchy. Asking the agency whether it would accept a claim the statistics could not support would have produced a documented negative that then sits in the file through review. |
| Type A after the Form 483 observation | Stage 4 | A Type A meeting is for a stalled program. One observation, closed within the response window, is not a stall. Requesting one would have signalled a significance the finding did not have. |
| Advisory committee request | Pre-NDA | A sponsor cannot convene one, and volunteering for one on a well-characterized mechanism in a crowded class invites public scrutiny of the tolerability profile without a corresponding benefit. |
That is the asymmetry that governs this table. A favorable answer helps the program; an unfavorable one is durable, discoverable and gets read by the review division that will later assess the submission. A question worth asking is one where the sponsor can act on either answer. A question asked in hope is a question that should have been answered internally first.
The corollary a program manager has to hold onto: a sponsor cannot make a review go faster, but can very easily make one go slower. Every one of the four meetings held changed something concrete. That is the standard the ones above failed to meet.
One genuine cost sits behind this discipline. Declining to ask means carrying an unresolved question at the sponsor's own risk, and the labeling outcome shows what that costs when the judgment is wrong. The comparative tolerability position was never tested with the agency because the internal assessment said it would fail — and it would have. But an internal assessment is not a record, and when Gate 6 reviewed the label finally granted, there was no documented agency position to point to, only a decision taken years earlier by people who had moved on. Not asking is a defensible choice; not writing down why you did not ask is not.