← Drug Development Suite NDA Submission · Vitalis Therapeutics Inc. · Illustrative end-state sample — forward-dated, not reconciled

Gate 5 Decision Record

Download Word
Program timeline · status 30 Sep 2028Illustrative end-state sample — forward-dated, not reconciled
Gate 0
Mar 2022
Go
Stage 1
Nonclinical
Gate 1
Dec 2022
Go
Stage 2
IND & Phase 1
Gate 2
Jun 2023
Go
Gate 3
May 2024
Go
Gate 4
Jun 2026
Go w/ conditions
Stage 4
Phase 3
You are here
Gate 5
Sep 2028
Gate 6
Oct 2029
Launch
Nov 2029

Vitalis Therapeutics Inc. — The recorded decision of the Development Committee at Gate 5, the NDA submission gate for VitaFlow (VTX-401), held 2028-09-30. Outcome GO (5-0). Includes the eight pivotal endpoint results, the must-meet and should-meet assessment, and the labeling discussion that changed the commercial case.

GO (5-0)
Outcome
6 / 6
Must-meets satisfied
3 / 5
Should-meets met
$15.8M
Tranche released
Contents
  1. Decision
  2. Results
  3. Must-Meet Assessment
  4. Should-Meet Assessment
  5. The Discussion That Mattered
  6. Authorization
  7. What This Record Establishes
  8. What Gate 5 Did Not Authorize
Forward-dated end-state sample. This record speaks from 2028-09-30 — two years after the 15 October 2026 status date the rest of the in-flight suite is written to. It is deliberately not reconciled against those artifacts.

The in-flight documents forecast; this one reports. Where they disagree, the disagreement is the point — it is what a governance record looks like when read forward rather than backward.

1. Decision

GateGate 5 — NDA Submission
Held2028-09-30
OutcomeGO (5-0)
Tranche released$15,800,000 — Stage 5, submission and review
NDA submitted2028-10-12
Filing accepted2028-12-11 — 60-day filing review completed
Action date set11 October 2029
The vote was unanimous and two endpoints missed their targets. Both belong in this record.

A gate outcome is not a verdict on whether everyone is pleased. It is a determination against criteria fixed before the data existed. Every must-meet was satisfied, so GO was available; two should-meets were not, so the record says so. A committee that had voted 5-0 and minuted only the successes would have produced a document nobody could use two years later.

2. Results

EndpointResultCriterionAssessment
Pivotal 301 — mean weight reduction, week 6814.2%≥15% target / ≥10% minbelow target, clears minimum
Pivotal 302 — mean weight reduction, week 6814.8%≥15% target / ≥10% minbelow target, clears minimum
Responder rate ≥5%, pooled81.3%≥80% targetmet
Responder rate ≥10%, pooled64.1%hierarchical secondarymet
Responder rate ≥15%, pooled41.7%hierarchical secondarymet
Waist circumference change−11.4 cmhierarchical secondarymet
GI-attributed discontinuation6.4%≤4% target / ≤7% minbelow target, inside minimum
CV sub-study — MACE hazard ratio0.91 (UB 1.18)UB <1.4met
Read the seventh row against the whole program. GI-attributed discontinuation came in at 6.4%, against a v3.0 target of ≤4% and a minimum of ≤7%.

That result cleared the bar that made the drug launchable and missed the bar that would have made it differentiated. Phase 2 had shown 5.1% at 36 weeks; the separation narrowed over 68 weeks at scale, which is the direction tolerability signals usually move and the reason the Target Product Profile recorded 24 weeks as insufficient evidence for a 68-week claim.

The commercial thesis for this asset was tolerability. The clinical result supports a product; it does not support the thesis.

The weight-reduction results tell a similar story more mildly. 14.2% and 14.8% are real, clinically meaningful reductions that comfortably clear the ≥10% minimum — and they sit below the ≥15% target that v2.0 of the TPP raised on the strength of Phase 1 data. Raising a target on early data is unusual, and this is the cost of it: a program that would have met its original target is recorded as missing its revised one.

3. Must-Meet Assessment

Assessed first, before any scoring, in the sequence set out in the Gate Decision Framework.

CriterionStatusEvidence
Both pivotal trials met the primary endpoint with statistical significancemetSuperiority to placebo demonstrated on percent weight reduction at week 68 in both 301 and 302.
Safety database adequate for the intended labelmet2,480 randomized, 68-week exposure, no unexpected serious safety signal. DMC recommended CONTINUE at every review.
CMC package complete and registration batches on stabilitymetAll three registration batches manufactured and on stability. GC-01 closed 2027-01-22.
Process validation completemetPPQ campaign completed Q2 2027, three consecutive conforming batches.
Agreed pediatric study plan in placemetiPSP agreed 2027-02; adolescent study deferred as a post-marketing requirement.
No unresolved matter that would prevent filing acceptancemetPre-NDA meeting 2028-08-10 confirmed content and format.

All six satisfied. GO was therefore available as an outcome — which is the only thing a must-meet assessment establishes. It does not make GO correct; it makes it possible.

GC-01 closed on 2027-01-22, fifteen months after the status date at which the in-flight suite records it as open. The purity method re-qualified, the third registration batch was manufactured and placed on stability, and PPQ started.

The Conditions Register refused partial closure on two of three methods. That refusal cost the program roughly four weeks of argument and is the reason this must-meet is recorded as met rather than as substantially met.

4. Should-Meet Assessment

CriterionStatusNote
Weight reduction at or above the target in the TPPnot met14.2% and 14.8% against a ≥15% target. Both clear the ≥10% minimum.
GI-attributed discontinuation at or below the targetnot met6.4% against a ≤4% target. Inside the ≤7% minimum.
Responder rates at or above targetmet81.3% at the ≥5% threshold.
Cardiovascular safety demonstratedmetMACE hazard ratio 0.91, upper bound 1.18, well inside the 1.4 requirement.
Market access evidence package completepartialGC-03 closed late, 2027-06. Advisory boards held, but the payer evidence was assembled after the pricing model had been built on it.
Three of five met, one partial, two not met — and the outcome was still GO.

That is the scoring model working as designed. Should-meets inform the decision; they do not control it. A program that missed two aspirational targets while satisfying every obligation required for a filing is a program that should file.

The alternative reading — that missing a differentiation target should have triggered RECYCLE — would have meant holding a completed pivotal program off the market to chase a tolerability margin that 68-week data had just demonstrated was not there. RECYCLE is for programs that can be fixed. This one could not be, and the Committee said so.

5. The Discussion That Mattered

The minuted debate was not about whether to file. It was about what to claim.

The comparative tolerability claim

J. Barrington (SVP Market Access) argued for pursuing a comparative tolerability statement in labeling on the strength of the 6.4% result. Dr. P. Raghunathan (SVP Regulatory) advised against, on three grounds:

The Committee accepted the regulatory position. No comparative tolerability claim was sought.

This is the moment the commercial case changed, and it was foreseeable from the moment the endpoint hierarchy was fixed.

The Protocol Summary records that argument being had at protocol finalization, more than two years earlier, and lost — because reordering the hierarchy to protect a commercial claim would have been visible to the agency as exactly that. The decision was correct then and its consequence arrived here.

A program that had promoted the differentiation endpoint would have had a stronger claim and a weaker filing. That trade was made once, deliberately, and this record is where the bill came due.

The abstention that wasn't repeated

M. Thornbury (CFO), who abstained at Gate 4 on un-refreshed gross-to-net assumptions, voted GO here. The commercial model had been rebuilt in Q1 2028 with current payer data. The revised net price assumption was $11,050 against the $12,500 the Gate 4 case had carried — a 12% reduction, and the reason the abstention had been recorded.

6. Authorization

ElementPosition
Tranche released$15,800,000
Cumulative released$217,000,000 — the full base authorization
Contingency remaining at this gate$9,890,000 of $26,040,000
Ceiling$243,040,000 — unchanged since Gate 4

Stage 5 covers submission, the review period, pre-approval inspection support and launch preparation to the action date. It is the smallest tranche in the program (7% of base) and the one with the least discretion attached — most of what happens in Stage 5 is responding to an agency on the agency's timetable.

Risk-adjusted value at this gate: $1,253.0M. Only the filing-to-approval conditional remains unresolved, at roughly 92%.

Against $15,800,000 of remaining authorization, that is the most favorable ratio in the program's history — and the correct reading is not that the program became a good investment at Gate 5. It is that the uncertainty which made it a poor one has been bought out, over five years, one tranche at a time. The expected value rose because the program spent money to resolve conditionals, exactly as the Charter §2 anticipated.

7. What This Record Establishes

What this record does not establish. It does not establish that the program will be approved — that is Gate 6. It does not establish that the commercial case holds; the 12% net-price revision and the late market-access evidence both point the other way, and the Post-Launch Review reports what came of that.

A gate record should be readable by someone who does not yet know how the story ends. This one is written that way on purpose.

8. What Gate 5 Did Not Authorize

A gate decision is bounded, and the boundary is as much a part of the record as the outcome. Gate 5 authorized two things: the release of the Stage 5 tranche of $15,800,000, and the submission of the NDA. It authorized nothing else, and four of the exclusions matter enough to record.

Not authorized by this gateWhy notWhere it is decided
ApprovalThe Committee has no authority over an agency decision and cannot vote one into existence. Everything after submission is the review clock running.FDA, target action date 11 Oct 2029.
LaunchCommercial launch is a separate authorization taken against an approval that did not exist on this date. Treating Gate 5 as launch authorization would commit the company to launch expenditure two years before there is a product to launch.Gate 6.
Commercial-scale manufacture for saleProcess validation is complete but the pre-approval inspection has not occurred. Product made now is launch inventory built at risk, not saleable goods.Approval, then release by Quality.
Promotional activityPre-approval promotion of an investigational product is prohibited, and the label being sought is not the label that will be granted. Disease-state education and payer value-dossier development continue; product claims do not.Approved labeling.
The most expensive misreading of a gate decision is treating authorization to proceed as confirmation of the outcome.

Gate 5 carried unanimously, 5-0, on a submission package the Committee judged complete. Nothing in that vote is a statement about whether VitaFlow would be approved, what label it would carry, or what it would be worth. The Committee assessed readiness to submit — a question about the last six months of work — and the record should not be read as answering any larger question, because it was not asked one.

The Gate 4 abstention is the counterpart to this. It recorded a reservation about a commercial assumption at the gate where that assumption was load-bearing. By Gate 5 the reservation was still live and still unresolved, and it was outside the scope of the question Gate 5 asked.

Two conditions were carried out of this gate into Stage 5, both verification-only: confirmation that the pre-approval inspection readiness plan is executed against the sites named in the submission, and a refreshed gross-to-net position tabled at Gate 6 rather than at the next quarterly review. Neither gated the submission. A condition that would have gated the submission would have produced a different outcome word than GO.