Sample · Public evidence only
K-1728: what should the first human study actually prove?
A worked Hive Prospect Decision Dossier showing how one early-development question can be reduced to the evidence that matters, the uncertainties that still control the decision, and the cheapest credible next truth.
01 · The decision
Design the first-in-human study to answer the portfolio question, not merely to complete Phase 1.
The public record supports an imminent Phase 1 start, a pan-mutant-selective PI3Kα thesis, and development interest in both cancer and vascular malformations. The decision is therefore not simply whether K-1728 can be dosed. It is what minimum human evidence would justify the next major allocation of capital and clarify which development path deserves priority.
02 · Evidence supporting the thesis
Why the program merits a decisive first study
- KymaThera says K-1728 is an oral, pan-mutant-selective PI3Kα inhibitor intended to spare wild-type PI3Kα function while retaining potency across major mutant classes.
- The company positions the program against known limitations of existing PI3Kα inhibitors, particularly tolerability constraints that can limit dosing and efficacy.
- KymaThera announced an $80 million Series B specifically to advance K-1728 into clinical development and said patient dosing is expected in Q4 2026.
Public sources: KymaThera corporate pipeline and Oct. 2026 financing announcement.
03 · What argues against overconfidence
What the public evidence does not yet establish
- Preclinical selectivity does not establish a clinically meaningful therapeutic window.
- Public materials do not yet establish the dose-response relationship needed to separate pharmacologic elegance from useful human activity.
- The relative value of oncology versus vascular-malformation development cannot be determined from the public record alone.
- No public evidence reviewed here establishes that one indication should dominate the initial capital-allocation decision.
04 · Critical unknowns
05 · Cheapest credible next experiment
A first study should be designed around a pre-agreed differentiation threshold.
06 · Alternative paths
| Path | What it proves | Advantage | Principal risk |
|---|---|---|---|
| Oncology-first | Human differentiation in a high-value mutant-driven cancer setting | Direct connection to the stated lead oncology thesis | Heterogeneity can obscure an early signal |
| Vascular-malformation-first | Target biology in a potentially more genetically concentrated setting | May provide a cleaner mechanistic read | Commercial and development path may be narrower |
| Parallel optionality | Preserves both paths until human PK/PD clarifies the winner | Maximizes information value | Consumes more capital and management attention |
07 · Go / modify / kill
08 · What I would do Monday morning
Three actions, in order.
- Write the differentiation threshold before dosing begins. Define exactly what PK, PD, tolerability, and early activity would constitute a meaningful win.
- Compare the first two indication paths apples-to-apples. Use actual protocol assumptions, patient availability, assay burden, site startup, and expected decision time rather than generic benchmark costs.
- Predefine the capital decision. State which result triggers oncology expansion, vascular-malformation expansion, parallel development, or stopping.
Why this is a sample, not a recommendation
This worked example uses only public information and deliberately preserves major unknowns. It does not substitute for company protocol data, unpublished preclinical evidence, regulatory interaction, site quotations, or management objectives. A paid dossier narrows the question with the buyer and works from the best available evidence for that specific decision.