One consolidated read on intratumoral immuno-oncology in melanoma: where oncology development has actually moved, how the intratumoral route and the immunostimulant class sit inside it, the response-rate and failure record after the first approval in the modality — and the private-financing, valuation and market-sizing picture underneath it, drawn straight from primary filings.
One consolidated, primary-source read on intratumoral immuno-oncology in melanoma: where development actually moved, the response and failure record after the first approval in the modality, and the capital picture underneath it — drawn from ClinicalTrials.gov, SEC filings and PitchBook.
Intratumoral innate-agonist programmes kept attracting acquirers, but companies ran out of money before the biology was tested. Disclosure in the innate / TLR-agonist cluster fell ~60–69%, the post–anti-PD-1 response band settled at ~20–25%, and the setting had no approved intratumoral therapy.
Interest without capitalThe first intratumoral therapy after anti-PD-1 won FDA approval on 6 Aug 2026 at 24.2% ORR, and the approved company’s own sizing is ~10,400 US patients a year. Capital has concentrated into fewer, larger, later rounds — while the exact intersection, intratumoral immunostimulants in melanoma, is just 41 trials in the entire registry, ~1 started since Jan 2024.
Validated, still near-emptyA freshly validated setting sitting on top of an emptied innate / intratumoral lane is the white space. KIN-112 — an intratumoral, non-viral TLR9/TLR7 innate agonist — enters exactly there. This is an implication of the data above, not a market forecast.
Outlook · implicationDevelopment trends by modality, the five-tier trial funnel (oncology → intratumoral immunostimulants), phase structure, the anti-PD-1-refractory response band, and the discontinuation record.
Comparable raises, market sizing, private-round valuations, acquisition comparables, and the active-investor picture in this exact niche.
Both cover intratumoral immunotherapy in melanoma — the Field is the IT / modality view, the Capital is the melanoma-market view.
Field and clinical-trial data on the intratumoral immuno-oncology space: disclosure trends by modality, the nested trial funnel, phase structure, the anti-PD-1-refractory response band, the failure record, the treatable melanoma population, and the preclinical science base.
A capital-markets read drawn from primary filings: the overall private-financing climate, market sizing, comparable raises and valuations, acquisition comparables, and who is actually active in this niche.
The data does not support "capital is flooding into the space." It supports something better, and more defensible.
Private biotech capital is contracting. Reg D offerings by US biotech and pharma issuers fell from $18.43B in the twelve months to June 2024 to $10.95B in the twelve months to June 2026 — down 41% — with offering counts down every year. A claim of a rising tide of capital would be the first thing to break under a partner's check.
What did happen in the last thirty days is far more valuable to us: on 6 August 2026 the FDA granted accelerated approval to TUDRIQEV (RP1) in combination with nivolumab, for exactly the population we are sizing — adults with unresectable advanced cutaneous melanoma who progressed on an anti-PD-1-based regimen. An FDA advisory committee had voted 10–3 a week earlier that the IGNYTE efficacy results were evaluable and clinically meaningful. That is the first regulatory validation of an intratumoral agent in anti-PD-1-refractory melanoma, and it arrived after two Complete Response Letters.
So the honest framing here: the regulatory path just got proven, the label just defined our market, and the private capital has not arrived yet. That is a timing argument, not an exit argument — and it is the only version of this the numbers actually carry.
Three things follow from this.
Disclosure intensity by modality, measured as the number of SEC filings mentioning each term per year. It is a proxy for where public-market attention and capital sit, and it separates cleanly into modalities that grew and modalities that contracted.
Four leading modality classes, shown for contrast — not a full list of immunotherapy modalities. KIN-112's own class, the innate / TLR-agonist cluster, is the one that fell (see Read below). SEC EDGAR full-text search, filings per year; 2026 covers 1 January to 2 September and is partial — read the shape, not the final point.
Indexed to 2019 = 100. Separating the delivery route from the mechanism it delivers.
Interventional trial counts. Each tier is a subset of the one above it. The right-hand column is the share of that tier's entire registry history that began in the last thirty-two months — a crude but consistent growth proxy.
Share of each tier's total registry history that started on or after 1 January 2024.
| Tier | Definition | Trials, all time | Started 2024+ | Share |
|---|
Phase buckets overlap where a study is registered as Phase 1/2 or 2/3, so buckets sum above the total. The ratio of early to late trials is the useful comparison, not the absolute counts.
Share of each tier's trials registered at Phase 3 or 4.
| Tier | Phase 1 | Phase 2 | Phase 3/4 | Total | Ph3-4 share | Ph1 : Ph3-4 |
|---|
Every reported objective response rate in anti-PD-1 refractory or resistant melanoma for intratumorally delivered or innate-agonist agents, with denominators attached.
Bar length is objective response rate; the number beside each bar is the evaluable denominator.
| Agent | ORR | n | Median DoR | Design | Note |
|---|
One figure requires particular care: a 70–78% response rate is sometimes cited for a TLR9 agonist in this class. That figure is from treatment-naive patients. The refractory figure for the same agent is 20.7%.
Every discontinued intratumoral immunostimulant programme that reached Phase 1/2 or beyond, separated by the reason development stopped.
| Agent | Route | Approved |
|---|
Every approved agent in this class is delivered locoregionally rather than systemically, and none is delivered by intratumoral injection.
We sized the US refractory anti-PD-1 melanoma population at roughly 10,000. Replimune's fiscal-2026 10-K, filed 29 June 2026, gives us that number with a company's BLA staked on it.
"We estimate approximately 13,000 patients progress on or after PD-1 treatment annually in the U.S. with approximately 80% of these patients eligible for treatment with RP1 in combination with nivolumab, if approved." Replimune Group, Form 10-K for fiscal year ended 31 March 2026 — filed 29 June 2026
13,000 × 80% = 10,400 addressable US patients per year. This is the strongest possible citation for the number: not an analyst estimate, but a company's own filed disclosure, made while its BLA was under FDA review, and now underwritten by an approval in that indication.
Site of care. Replimune states these patients are treated in the outpatient setting, do not require hospitalisation, and are almost all in settings with interventional radiology onsite or readily accessible. That is a filed answer to the "who will actually inject this" objection.
Incidence direction. The 10-K notes melanoma incidence has risen steadily for thirty years. Our denominator grows.
IGNYTE registrational cohort, anti-PD-1-failed cutaneous melanoma, n = 140: confirmed ORR 32.9% (15.0% complete response), median duration of response 33.7 months, median overall survival 32.9 months. Survival 75.3% at one year, 63.3% at two, 47.8% at three — 83.5% among responders.
Population difficulty, for like-for-like comparison: 65.7% primary anti-PD-1 resistance, 56.4% PD-L1 negative, 46.4% prior anti-CTLA-4. Published J Clin Oncol 43(33):3589–3599.
US figures. Each step is sourced separately; the final step is a derivation, not a reported figure.
US approvals. Only two have ever been delivered intratumorally, eleven years apart.
Every Reg D offering by a US biotech or pharmaceutical issuer, 2022 Q1 – 2026 Q2, de-duplicated to the offering level using the SEC file number so amendments are not double-counted.
A pre-IND or Phase 1 ask is not competing for the capital that disappeared. The $5–50M bands lost roughly $0.85B of annual volume across two years; the ≥$100M band lost $4.95B. The scarcity is at crossover and pre-IPO stage, not at our stage.
Use this to pre-empt the "biotech funding is dead" objection rather than avoiding it — the objection is true about a part of the market we are not in.
Filtering the same Form D universe to companies working on intratumoral delivery, oncolytic virus, or innate-immune agonism: 25 offerings, $152.1M disclosed, median $1.9M across the 24 months to Sept 2026. Nearly all of it is small financing by already-public micro-caps, not venture rounds.
Broadening to the full melanoma / immuno-oncology cohort: 45 offerings, 29 issuers, $2.80B over the same window — but that number is dominated by large platform and T-cell-engager companies, not by anything intratumoral.
| Window | Offerings | Issuers | Total sold | Median | Mean | Rounds ≥$100M | Top-decile share |
|---|---|---|---|---|---|---|---|
| Jul 2022 – Jun 2023 | 970 | 832 | $15.09B | $3.40M | $15.56M | 32 | 64% |
| Jul 2023 – Jun 2024 | 903 | 752 | $18.43B | $2.68M | $20.41M | 53 | 68% |
| Jul 2024 – Jun 2025 | 798 | 682 | $13.95B | $3.28M | $17.48M | 34 | 71% |
| Jul 2025 – Jun 2026 | 767 | 646 | $10.95B | $2.53M | $14.28M | 23 | 69% |
Private financing measured two ways: from Reg D filings, which capture US private placements exhaustively but carry no round label or valuation, and from PitchBook, which carries both but covers a curated universe. The two are complementary and are not additive.
Oncology and biotechnology venture, all stages, global. Source: PitchBook. 2026 covers 1 January to 2 September and is partial.
Median values across the same PitchBook universe. Pre-money and post-money are reported medians, not derived.
Private financings of companies with an active Phase 1 or early Phase 1 programme. Source: PitchBook. 2026 is year to date.
Median round size, pre-money and post-money valuation by labelled series, oncology and biotechnology venture. Source: PitchBook.
Total amount sold, US biotechnology and pharmaceutical issuers, de-duplicated to the offering level. Source: SEC Form D structured data sets.
Capital by round-size band across four rolling twelve-month windows. Bands are ordered magnitude, so the ramp runs light to dark.
SEC Form D offerings of $2M or more, July 2024 to June 2026, with company age taken from self-reported year of incorporation.
Immuno-oncology acquisitions, 2022 to 2026, by development stage at announcement. Source: PitchBook.
Disclosed private rounds, from PitchBook. Pre-money is shown where reported.
| Company | Round | Amount | Date | Focus |
|---|
The label-based benchmark underlying Figure 13. Source: PitchBook.
| Series | Deals | Median round | Median pre-money | Median post-money |
|---|
Peer-reviewed publication counts per year from NCBI PubMed E-utilities, filtered to animal, in vivo and preclinical studies. Exact result counts, not estimates.
Both series indexed to 2019 = 100. Publications on the left axis convention, filings on the same index.
| Research area | 2019 | 2021 | 2023 | 2025 | Change |
|---|
Form D does not record financing stage, so stage is proxied by company age at the round: year of incorporation as reported by the issuer, subtracted from the filing date. Filtered to rounds of $2M or more, which removes friends-and-family and SPV noise.
| Stage proxy | n | 25th | Median | 75th | 90th |
|---|---|---|---|---|---|
| Age 0–1y — seed / newco | 122 | $4.0M | $6.0M | $14.7M | $39.8M |
| Age 2–3y — Series A / IND-enabling | 113 | $5.0M | $10.0M | $25.0M | $46.8M |
| Age 4–6y — Series B / clinical | 131 | $4.6M | $10.0M | $30.4M | $77.0M |
| Age 0–3y combined — our band | 235 | $4.4M | $7.4M | $18.8M | $44.3M |
Form D discloses the amount sold, not the pre-money. Private-round valuations are not in the SEC record at all, so any "valuations at that stage" figure cannot be sourced to sec.gov and should not claim to be. What we can defensibly anchor to is round size, which is what the table above gives, and the public comparable: Replimune at approval, with one accelerated-approval product and a Phase 3 running, carries an implied equity value of roughly $1.32B at its 10 August 2026 offering price.
Two adjacent Form D data points we can cite for what large early rounds look like right now: Candid Therapeutics raised $206.2M in September 2024 as a company incorporated that same year, and Arsenal Biosciences raised $325.4M the week before. Both are the exception, and both are platform companies rather than single-asset — useful for showing what the top of the market funds, not as a benchmark for our ask.
The two comparables Danny wants rebuilt as a tailwinds story. Both now carry a validation event dated inside the last eight months.
Replimune took eleven years and roughly $1.28B to put the first intratumoral oncolytic into the anti-PD-1-refractory melanoma label. It absorbed two Complete Response Letters — July 2025 and April 2026 — and still got there, on accelerated approval, after an advisory committee voted 10–3 in its favour. The market then handed it $150M four days after approval.
The tailwind claim we can make from this and defend: the FDA has now accepted intratumoral oncolytic immunotherapy plus checkpoint blockade as an approvable regimen in this exact population. Every subsequent entrant inherits a defined endpoint, a defined comparator problem, and a labelled patient population. That is a regulatory precedent, which is worth more to a Series A than an exit multiple.
| Date | Instrument | Amount | Detail |
|---|---|---|---|
| Sep 2015 | Series A preferred | $15.0M | Omega, Forbion, Atlas Venture |
| Mar 2017 | Series A, 2nd close | $15.0M | Same syndicate |
| Jul–Sep 2017 | Series B preferred | $55.0M | Two closings at $63.79 |
| Jul 2018 | IPO | $103.3M | Net; $15.00 per share |
| Nov 2019 | Follow-on | $85.6M | Net |
| FY2020 | Venture debt | $10.0M | Drawn |
| Jun 2020 | Follow-on | $107.8M | Net |
| Oct 2020 | Follow-on | $270.0M | Net |
| Dec 2022 | Follow-on | $242.6M | Net |
| FY2023 | Debt | $28.2M | Drawn |
| FY2024 | Debt | $15.0M | Drawn |
| FY2025 | Follow-on | $155.7M | Net; Nov 2024 offering |
| FY2026 | Debt | $35.0M | Drawn |
| Aug 2026 | Follow-on | $140.5M | Net; $150.0M gross at $12.06 |
| Total | Equity + debt | ~$1.28B | Accumulated deficit $1.332B at 30 Jun 2026 |
The economics, verbatim from Moderna's 2025 10-K: Merck paid a $200 million upfront under the personalised cancer vaccine collaboration; in September 2022 Merck exercised its option and in October 2022 paid a $250 million option exercise fee. Costs and any profits or losses are shared equally worldwide.
What that bought: a randomised Phase 2b in adjuvant high-risk resected melanoma that reduced risk of recurrence or death by 44% (HR 0.56, 95% CI 0.31–1.02, one-sided p = 0.0266) against pembrolizumab alone — the first efficacy demonstration for an mRNA cancer treatment in a randomised melanoma trial. Breakthrough Therapy designation followed in February 2023.
The tailwind datum here is dated January 2026: Moderna and Merck reported five-year Phase 2b data showing sustained recurrence-free survival benefit in stage III/IV resected melanoma. Eight Phase 2 and Phase 3 trials are now running across tumour types. A top-five pharma is funding half of a melanoma immunotherapy programme through Phase 3 and past five-year follow-up — that is the validation claim, and it needs no exit multiple attached.
The exit-path framing is weak, and the data makes the case: presented as an exit path these numbers are weak. Presented as evidence of what large acquirers will pay for pre-data intratumoral assets, they are strong.
Regeneron / Checkmate Pharmaceuticals, April 2022. Cash tender at $10.50 per share for a company whose sole product candidate was vidutolimod (CMP-001) — a TLR9 agonist delivered intratumorally, in melanoma, non-melanoma skin cancer and head and neck cancer. 22,031,231 shares outstanding, so roughly $231M of equity value and about $250M fully diluted.
The detail that matters: four months before the offer, Checkmate had pushed its anti-PD-1-refractory melanoma data readout out to the second half of 2023. Regeneron bought a non-viral intratumoral innate-immune agonist in this exact population, at Phase 2, before the pivotal data existed. That is the closest structural analogue to our asset that the SEC record contains.
XOMA Royalty / Turnstone Biologics, July–August 2025. Tender offer at $0.34 per share plus a contingent value right. Turnstone was the leading oncolytic-virus-plus-TIL company; it IPO'd in 2023 and was acquired as a wind-down after discontinuing its lead programme.
Worth stating plainly. A room that knows the field already knows this deal, and volunteering it buys credibility for the Checkmate comparison. The distinction to draw is that Turnstone failed on its asset, not on the modality or the regulatory path — and the TUDRIQEV approval one year later is the evidence.
| Target | Acquirer | Announced | Modality | Stage at acquisition | Consideration | Signal |
|---|---|---|---|---|---|---|
| Checkmate Pharmaceuticals vidutolimod / CMP-001 |
Regeneron | 19 Apr 2022 | TLR9 agonist, intratumoral, non-viral | Phase 2 — refractory melanoma readout guided to 2H 2023 | $10.50/sh ~$250M |
Validating |
| Turnstone Biologics TIDAL-01 |
XOMA Royalty | 11 Jul 2025 | Oncolytic virus + tumour-infiltrating lymphocyte | Wind-down after lead programme discontinued | $0.34/sh + CVR |
Cautionary |
A full-text search of every Schedule 14D-9 and merger proxy filed between January 2023 and September 2026 returns exactly one transaction mentioning "oncolytic" — the Turnstone wind-down. There has been no acquisition of a healthy intratumoral oncology company in three and a half years.
Any implication of a live M&A market in this niche is the claim that breaks under diligence. What the record does support: a proven regulatory route, a labelled population, a big-pharma-funded melanoma programme in Phase 3, and one pre-data acquisition at ~$250M by a top-tier acquirer. The four that hold read as validation rather than exit.
Derived from Schedule 13D and 13G filings, 2024–2026, across fifteen melanoma and immuno-oncology issuers. Disclosed 5%-plus positions — real holders rather than reported interest. The market-structure charts below are shown in both modes; the holder-level target list is internal.
Investment count over the twenty-four months to September 2026, oncology and biotechnology venture. Source: PitchBook.
Investors participating in melanoma private financings, 2024 to 2026, by number of deals. Source: PitchBook melanoma indication screen.
Wellington Management — Immatics, Immunocore
Baker Bros. Advisors — Replimune, Immatics, Immunocore
T. Rowe Price — Replimune, Immunocore
Perceptive Advisors — Immatics, Iovance
Fidelity / FMR — Nuvation, Turnstone
Redmile, RTW, Rock Springs, Stonepine, Boxer Capital, Avoro, HBM Healthcare, Point72, Quogue, MHR — single names across the cohort
Novo Holdings — IO Biotech (three filings, the most active single-name venture holder in the cohort)
Omega Fund and Forbion — Replimune, from the 2015 Series A and still disclosed
Venrock Healthcare, Vivo Capital, Curative Ventures — Instil Bio
Decheng Capital — CG Oncology
General Atlantic — Immunocore
Clal Biotechnology, GKCC — Elicio Therapeutics; ARYA Sciences, Athos KG — Immatics
Omega Funds and Forbion seeded Replimune's Series A in September 2015 and are still on the register eleven years later, through two CRLs and an approval. They are the only investors in the SEC record who have taken an intratumoral oncolytic from newco to label. They know the CMC and tox burden we are about to describe, because they funded it.
Novo Holdings is the most persistently active venture holder in melanoma specifically, via IO Biotech.
Beyond those, the specialist crossover set — Baker Bros, Wellington, Perceptive, RTW, Redmile — is where the melanoma-literate money sits, but those are Series B-and-later relationships. Start them as relationship-building rather than as this round's target.
So that any number here survives a partner pulling the filing.
Eighteen quarterly Form D structured data sets (2022 Q1 – 2026 Q2) were downloaded from sec.gov and joined across the submission, offering and issuer tables — 259,042 filings, of which 4,603 are Biotechnology or Pharmaceuticals issuers.
Filings were collapsed to offerings by grouping on issuer CIK plus SEC file number, so an original Form D and its amendments count once. Each offering is dated to its first filing and valued at the maximum amount sold reported across all its filings.
Recent quarters understate. Amounts rise as amendments are filed, so 2026 Q1–Q2 will be revised upward.
No stage, no valuation. Form D records neither. Stage is proxied by company age; valuation is simply absent from the SEC record.
Coverage. Form D captures Reg D exempt offerings only — it includes PIPEs by listed micro-caps and excludes non-US rounds and offerings under other exemptions. The intratumoral cohort figures are therefore a floor, not a census.
Cohort construction. Company universe was built from EDGAR full-text search on the field's own vocabulary, then name-matched against Form D. Matching by name will miss private companies whose filings do not use these terms.
| Claim | Filer | Form | Filed | Accession |
|---|---|---|---|---|
| TUDRIQEV accelerated approval in anti-PD-1-progressed melanoma | Replimune Group | 8-K | 7 Aug 2026 | 0001104659-26-092246 |
| Advisory committee votes 10–3 that IGNYTE results are meaningful | Replimune Group | 8-K | 31 Jul 2026 | 0001104659-26-088857 |
| Complete Response Letter #1 | Replimune Group | 8-K | 22 Jul 2025 | 0001104659-25-069489 |
| Complete Response Letter #2 | Replimune Group | 8-K | 13 Apr 2026 | 0001104659-26-042141 |
| 13,000 patients / 80% eligible; IGNYTE efficacy; CRL history | Replimune Group | 10-K FY2026 | 29 Jun 2026 | 0001628280-26-045886 |
| $150.0M offering at $12.06; 82,716,923 shares outstanding | Replimune Group | 424B5 | 10 Aug 2026 | 0001104659-26-093040 |
| Series A and Series B terms; IPO at $15.00 | Replimune Group | 424B4 | 23 Jul 2018 | 0001047469-18-005126 |
| Merck $200M upfront, $250M option fee, 50/50 share; Phase 2b HR 0.56 | Moderna | 10-K FY2025 | 20 Feb 2026 | 0001682852-26-000033 |
| Regeneron tender at $10.50; 22,031,231 shares; CMP-001 at Phase 2 | Checkmate Pharmaceuticals | SC 14D-9 | 2 May 2022 | 0001193125-22-135583 |
| XOMA tender at $0.34 plus CVR | Turnstone Biologics | SC 14D-9 | 11 Jul 2025 | 0001193125-25-157832 |
| $206.2M raise by a company incorporated the same year | Candid Therapeutics | Form D | 13 Sep 2024 | 0002035778-24-000001 |
| $325.4M raise | Arsenal Biosciences | Form D | 4 Sep 2024 | 0001945638-24-000001 |
Our one-screen read on intratumoral immuno-oncology in melanoma — where we were, what is happening now, and where it points. Every number here is drawn from the figures on the Field and Capital tabs.
Everything on this site is built from two primary-source workbooks. Both are downloadable below; every figure traces back to a sheet in one of them.
The source data behind The Field. Twelve sheets: oncology direction, the modality filing-trend, the immunostimulant and intratumoral tiers, the intratumoral-immunostimulant intersection, the preclinical pipeline and research, the glatiramoid white-space, and melanoma epidemiology and approved therapies.
⇩ Download SEC / EDGAR packBuilt from primary records on sec.gov and ClinicalTrials.gov, retrieved 4 September 2026 — every Field figure traces to a sheet here.
The source data behind The Capital. Kinimmune’s own PitchBook profile, melanoma and oncology Phase 1 comparables, the field funnel and phase matrix, and the SEC-fact reconciliation.
⇩ Download PitchBook packRetrieved 4 September 2026 — every Capital figure traces to a sheet here. PitchBook data is licensed; internal use only.
Reed Jobs’ oncology-only fund is uniquely positioned for KIN-112: they already own a tumor-ablation asset, run a grant-to-venture funnel we can ride, and back exactly the vaccine/adjuvant space our innate agonist sits beside — here is the intelligence and the play.
Yosemite was founded in August 2023 by Reed Jobs and spun out of Emerson Collective, the organization of his mother Laurene Powell Jobs, where Reed had served as Managing Director of Health for roughly eight years. [Forbes Aug 2023; STAT Dec 2023; Wikipedia]
The firm is named after Yosemite National Park, where Steve Jobs and Laurene Powell Jobs married. Its mission is to make cancer non-lethal within our lifetime. It is oncology-only, based in San Francisco, staffed by roughly 10 professionals at launch (a team of 17 by mid-2026), and invests across all modalities. [BioPharma Dive; TechCrunch Jul 2026] By July 2026 the portfolio is close to 25 companies across both funds, with two scientific failures acknowledged. [TechCrunch 11 Jul 2026]
Yosemite runs a for-profit venture fund alongside a philanthropic donor-advised grant arm — the two halves feed each other, which is central to how we get in the door.
Two funds in ~2.5 years: launch and Fund I ($400M) in 2023, Fund II ($350M offering) by January 2026.
Sources · SEC EDGAR Form D (CIK 1970826, 2092713); Forbes; BioPharma Dive
Primary source (SEC EDGAR, Form D): Yosemite Fund I, L.P. (Delaware) filed its Form D on 27 July 2023 with a $400M offering; Yosemite Fund II, L.P. (Delaware) filed its Form D on 29 January 2026 with a $350M offering. [SEC EDGAR: CIK 1970826, CIK 2092713]
Offering figure (SEC, primary). Press framed it as ~$200M raised / a $263M first fund — amounts closed vs the offering target.
Offering figure (SEC, primary).
Conflict: internal PitchBook pack says ~$1.2B; press says “>$1B.” Includes capital managed for hospitals & endowments.
Show the SEC offering figures as primary; note the press framing of closed amounts as secondary.
First close >$200M announced 29–30 January 2026 against the $350M target (the same $350M as the SEC Form D offering). LPs named in press: Amgen, Memorial Sloan Kettering, MIT and John Doerr. AUM is now reported at >$1B. No final close has been reported as of September 2026 — they are placing fresh Fund II capital during our October window. [Forbes 29 Jan 2026; BioPharma Dive; Venture Capital Journal; Endpoints]
Hybrid for-profit + nonprofit. Roughly one-third of the fund goes to companies Yosemite builds from scratch internally.
SEC Form D offering figures: Fund I $400M (Jul 2023), Fund II $350M (Jan 2026).
Sources · SEC EDGAR Form D (CIK 1970826, 2092713)
Median deal size rose roughly 10× in three years — $9.1M to $91M.
Sources · Internal PitchBook pack
Roughly one-third of the fund backs companies Yosemite builds from scratch internally.
Sources · Internal PitchBook pack
Mechanics: 2.5% of the fund flows into a donor-advised fund (nonprofit), plus roughly $1M/year from management fees. Grants are “no strings, no IP taken.” [TechCrunch 2023]
The ACS–Yosemite program is a $13.2M commitment (announced Mar 2024, $330K/grant), run in two cycles so far: 2024 (year 1) — 20 awardees, >$6M, themes immuno-oncology / cell therapy + AI; and 2025 (year 2) — 19 awardees, $6.27M, themes cancer vaccines + targeted toxins. With care-delivery grants, >$18M total is deployed. [pressroom.cancer.org; cancerletter.com]
Across grant cycles. [PRNewswire]
19 grants × $330,000, in two themes. [ACS cancer.org]
An academic grant de-risks a discovery in the university lab; the scientist then returns to Yosemite for startup capital. The grant program is a feeder into the venture fund.
Proof it works: Doudna’s lab grant → Azalea Therapeutics; Craig Crews (Yale) grant → seeded Quarry Thera; Jeremiah Johnson (MIT) → published in Nature Biotechnology.
Care-delivery grants: Mayo Clinic, City of Hope. BrightEdge is ACS’s VC arm and runs the donor-funded ACS Impact Venture Fund (AIVF).
The 2026 ACS–Yosemite Award changed themes to (1) non-genetic regulation in cancer (alternative splicing, RNA modifications, non-canonical translation, post-translational modifications) and (2) induced proximity / protein modulation beyond degradation. Same money: up to $300K direct + 10% indirect = $330K total, two years, non-renewable. Window 16 March – 24 June 2026 (closed), peer review September, notification November 2026, grants start 1 January 2027. [cancer.org RFA]
Applicants must hold a full-time faculty post at one of 33 invited institutions (MIT, Stanford, Harvard, Yale, UCSF, UC Berkeley, MSK, MD Anderson, Dana-Farber-affiliated Harvard, Johns Hopkins, Penn, Duke, WashU, UNC, UT Southwestern, Mayo, City of Hope, Fred Hutch, Scripps, Rockefeller, Caltech, Columbia, Cornell, Dartmouth, Mount Sinai, IPD/UW, and abroad Oxford, Cambridge, UCL, Imperial, ETH Zurich, NKI, Peter MacCallum, A*STAR). Washington University in St. Louis — Cory Berkland’s institution since 2023 — is on the list. So the door is not closed by eligibility: it is closed for 2026 only because the window ended on 24 June and the themes are off-thesis for an innate agonist. Cory’s 2024 award is proof of prior selection, and the 2027 cycle (window opens ~March 2027) is a real re-entry via WashU if the themes fit. [cancer.org RFA, retrieved Sept 2026; WashU McKelvey]
2025-cycle awards now under way (ACS Spring 2026 list): Carolyn Bertozzi (Stanford) — lysosomally-targeted ADC for renal cell carcinoma; Kai Wucherpfennig (Dana-Farber) — enhancing tumor infiltration by vaccine-induced T cells; Xin Zhou (Dana-Farber) — potentiating ADC activity by controlled co-engagement. [cancer.org, Spring 2026 new-grants list]
Full grant table — all Yosemite–ACS awards, 2024 & 2025 · 39 × $330K ≈ $12.9M research (of >$18M total incl. care-delivery)
| Year | Awardee | Institution | Theme | Research focus |
|---|---|---|---|---|
| 2024 | Cory Berkland | Washington University in St. Louis | IO / cell therapy + AI | Engineering GM-CSF as tumor immunotherapy (MP-GMCSF) |
| 2024 | Juan Cubillos-Ruiz | Weill Cornell | IO / cell therapy + AI | T-cell cytoskeletal integrity for immunotherapy |
| 2024 | Gautam Dantas | Washington University in St. Louis | IO / cell therapy + AI | Yeast as oral delivery for immunomodulators |
| 2024 | Eric Duncavage | Washington University in St. Louis | IO / cell therapy + AI | ML to identify dysplasia in blood/marrow |
| 2024 | Yi Fan | University of Pennsylvania | IO / cell therapy + AI | Reversing immunosuppressive vasculature |
| 2024 | Christopher Gibbons | MD Anderson | IO / cell therapy + AI | Electronic patient-reported-outcome feedback |
| 2024 | Saar Gill | University of Pennsylvania | IO / cell therapy + AI | Inhibiting suppressive myeloid cells |
| 2024 | Brian Grindel | MD Anderson | IO / cell therapy + AI | Re-activating anti-tumor immunity |
| 2024 | Jeremiah Johnson | MIT | IO / cell therapy + AI | Antibody-bottlebrush prodrug conjugates |
| 2024 | Roarke Kamber | UCSF | IO / cell therapy + AI | Chimeric macrophage receptors for phagocytosis |
| 2024 | Kenneth Kehl | Dana-Farber | IO / cell therapy + AI | Open-source AI clinical-trial matching |
| 2024 | Dan Landau | Weill Cornell | IO / cell therapy + AI | Innate immunity for IO therapeutics |
| 2024 | Mark Leick | Massachusetts General Hospital | IO / cell therapy + AI | VEGF-blocking CAR-T for solid tumors |
| 2024 | Daniel Marcus | Washington University in St. Louis | IO / cell therapy + AI | AI multidisciplinary tumor-board platform |
| 2024 | Matthew Porteus | Stanford | IO / cell therapy + AI | Genome editing for stem-cell immunotherapy |
| 2024 | Anthony Rongvaux | Fred Hutchinson | IO / cell therapy + AI | T-cell infiltration mechanisms |
| 2024 | Debattama Sen | Massachusetts General Hospital | IO / cell therapy + AI | Epigenetic reprogramming of cell therapy |
| 2024 | George Souroullas | Washington University in St. Louis | IO / cell therapy + AI | Chromatin accessibility in tumor immunity |
| 2024 | Edus Warren | Fred Hutchinson | IO / cell therapy + AI | LLMs for T-cell antigen recognition |
| 2024 | Eric Winer | Yale | IO / cell therapy + AI | AI to reduce breast-cancer overtreatment |
| 2025 | Steven Elledge | Harvard Medical School | Cancer vaccines | Dark-proteome epitopes in cancer cells |
| 2025 | William Freed-Pastor | Dana-Farber | Cancer vaccines | Cryptic-antigen vaccines for pancreatic cancer |
| 2025 | William Kaelin | Harvard Medical School | Cancer vaccines | Endogenous retroviruses as vaccine targets |
| 2025 | Mark Lee | Yale | Cancer vaccines | Non-canonical tumor antigens |
| 2025 | Alexander Marson | UCSF | Cancer vaccines | Cancer-antigen immunogenicity atlas |
| 2025 | Alexander Stegh | Washington University in St. Louis | Cancer vaccines | Glycolipid spherical-nucleic-acid vaccine platform |
| 2025 | Eric Thompson | Duke | Cancer vaccines | NK-cell immunogenicity of peptide vaccines |
| 2025 | Catherine Wu | Dana-Farber | Cancer vaccines | Shared noncanonical antigens, off-the-shelf vaccines |
| 2025 | Kai Wucherpfennig | Dana-Farber | Cancer vaccines | Enhancing tumor infiltration by vaccine T cells |
| 2025 | Carolyn Bertozzi | Stanford | Targeted toxins | Lysosome-targeted ADCs for renal cell carcinoma |
| 2025 | Jennifer Doudna | UC Berkeley | Targeted toxins | Cas12a2 for targeted tumor elimination |
| 2025 | Michael Elowitz | Caltech | Targeted toxins | Engineered protein circuits for cancer therapy |
| 2025 | Possu Huang | Stanford | Targeted toxins | Immunotherapy targeting KRAS mutants |
| 2025 | Michael Kharas | Memorial Sloan Kettering | Targeted toxins | Degrader-antibody conjugates |
| 2025 | Funda Meric-Bernstam | MD Anderson | Targeted toxins | Trastuzumab-resistance mechanisms |
| 2025 | William Sellers | Broad Institute | Targeted toxins | Surfaceome targets for biliary-tract cancer |
| 2025 | Jamie Spangler | Johns Hopkins | Targeted toxins | Multispecific downregulating antibodies |
| 2025 | James Wells | UCSF | Targeted toxins | Extracellular protein degradation for ADCs |
| 2025 | Xin Zhou | Dana-Farber | Targeted toxins | ADC potentiation via receptor co-engagement |
Two themes, 19 awards × $330K = $6.27M in the 2025 (year-2) cycle.
Sources · cancer.org (ACS–Yosemite 2025 grants)
The 2025 cycle ($6.27M) is roughly a third of the >$18M deployed across all cycles.
Sources · PRNewswire; cancer.org
Named companies (~11 of ~16–25 total). Roles per the internal PitchBook pack; PB-vs-press conflicts flagged.
| Company | Focus | Deal & role |
|---|---|---|
| Chai Discovery Win | AI protein/drug design | Series C $400M Jun/Jul 2026 (Index, Kleiner, Sequoia); earlier $130M @ $1.3B; total >$600M; used by Lilly/Pfizer/Novartis. PB $130M vs press $400M |
| Tune Therapeutics Win | Epigenetic editing / gene therapy | ~$175M, co-led NEA / Yosemite / Regeneron Ventures / Hevolution |
| Solve Therapeutics Lead | ADCs for solid tumors | Yosemite LED $120M Nov 2025. PB $321M vs press $120M |
| HistoSonics ★ Win | Tumor ablation / histotripsy (device) | Growth round, $3.75B val, 22 Jun 2026, Reed Jobs |
| Shinobi Therapeutics | iPS-derived T-cell therapy | Series A-II 6 Aug 2024; total $119M (participant) |
| Quarry / Quarry Thera | Induced proximity / targeted protein degradation | ~$32M Series A1; grant-seeded (backer) |
| Azalea Therapeutics | Doudna-lab spinout, Cas12a2 | $82M total (key investor) |
| Fourier / Fourier Health | Oncology / AI | ~$8.4M seed (PB; press unconfirmed) |
| Braveheart Bio ⚠ off-thesis | Cardiac myosin inhibitor (off-oncology) | $185M launch 2025 (Third Rock / RA Capital / Sozo) |
| Lomond Therapeutics Holdings | Public biotech | Yosemite Fund I disclosed >5% stake via SEC Schedule 13G, Jan 2025 |
| Turquoise | San Diego, founded 2020 | $40M Series C, 17 Mar 2026 — Yosemite participated (Tracxn / CB Insights) |
Counts vary: Tracxn 16, CB Insights 9, press ~20–25 total; ~⅓ are internal build-from-scratch; 2 have failed for scientific reasons; the complete list is PitchBook-only.
HistoSonics ($3.75B, grey) is a valuation, not a round; Solve ($120M, orange) is the deal Yosemite led. Log scale spans $8.4M to $3.75B.
Sources · Internal PitchBook pack; BusinessWire; Cooley; NCBiotech; SEC EDGAR
Named companies span the full modality map — AI, editing, ADC, cell therapy, devices, degraders — with Braveheart the lone off-thesis (cardiac) name (grey).
Sources · Internal PitchBook pack (named companies only)
[BioPharma Dive; Traded VC; Endpoints]
Yosemite syndicates widely — Chai drew four named co-investors (GC, Index, Kleiner, Sequoia); ACS BrightEdge recurs as a co-investor and our on-ramp.
Sources · BioPharma Dive; Traded VC; Endpoints; company releases
HistoSonics: Yosemite already owns a tumor-ablation position. KIN-112 is the drug given during that procedure — a portfolio-level argument unique to Yosemite.
The ACS door — 2026 reality: Cory Berkland (Washington University in St. Louis) is a 2024 ACS–Yosemite grantee, so we are already in their funnel, and ACS BrightEdge is a co-investor. But the 2026 cycle is closed (deadline 24 June), its themes are non-genetic regulation and induced proximity — off-thesis for an innate agonist — so the grant path cannot be re-entered before the 2027 cycle. Cory’s institution, Washington University in St. Louis, is among the 33 invited institutions, so the door reopens for us in 2027. Until then it is a credibility signal, not a live door.
Reed’s own rules (TechCrunch, 11 Jul 2026): “CVs removed from consideration”; founders are told to email him directly; the 2026 thesis is undruggable targets (p53 via three companies, KRAS, Myc, β-catenin) plus AI, across modalities that explicitly include immunotherapy. So KIN-112 should be pitched as immunotherapy given during ablation (the HistoSonics fit), not as a target play.
Our $7M ask vs their ~$51M median is an order of magnitude low. Present the $7M as the tranche to IND, and name the Series A they’d lead at IND — two doors — inviting them to price the next round.
~20–30% for a direct lead of the $7M as-is; meaningfully higher if reframed.
The $7M ask is an order of magnitude below the ~$51M median — reframe it as the tranche-to-IND plus a named Series A they’d lead.
Sources · Internal PitchBook pack (median round)
~20–30% for a direct lead of the $7M as-is — meaningfully higher if reframed.
Sources · Internal assessment