Oncology and CNS Therapeutics

Technical evidence

1. Granted patent: anti-PD-1 antibody composition and methods of use

A granted U.S. patent — not an application — records enforceable claims for anti-PD-1 antibodies and methods of treating cancer:

US 8,952,139 B2, “Human monoclonal antibodies to programmed death 1 (PD-1) and methods for treating cancer using anti-PD-1 antibodies alone or in combination with other immunotherapeutics,” granted 2015-02-10. Accessed 2026-05-09.

patents.google.com

This is a valid Tier 1 patent source: a USPTO record via Google Patents. It is evidence that anti-PD-1 antibody composition and treatment claims are recorded in the patent system. It is not proof that any specific commercial product works or is commercially viable. It also does not, by itself, prove that this exact patent protects pembrolizumab (Keytruda); Keytruda’s specific patent estate is broader and requires assignment and Orange Book records for product-level patent mapping.

2. Peer-reviewed clinical evidence: KEYNOTE-006

The strongest efficacy evidence for the checkpoint mechanism in advanced melanoma is a peer-reviewed Phase 3 trial, not a preprint:

Robert C, et al. “Pembrolizumab versus ipilimumab in advanced melanoma.” N Engl J Med 2015;372:2521-2532. DOI: 10.1056/NEJMoa1503093. Published 2015-06-25. Accessed 2026-05-09.

nejm.org

KEYNOTE-006 randomized 834 patients with advanced melanoma to pembrolizumab every 2 weeks, pembrolizumab every 3 weeks, or ipilimumab. The paper reports a 12-month overall survival estimate of 74.1% for pembrolizumab every 2 weeks versus 58.2% for ipilimumab [SINGLE-SOURCE] — these are trial-specific Kaplan-Meier estimates. This is a Tier 1 peer-reviewed source with a verifiable DOI.

3. CNS tumor evidence: CheckMate 143 exploratory cohorts

CNS glioblastoma is the key bridge to “CNS therapeutics” in this note. The relevant peer-reviewed, DOI-verified paper is:

Omuro A, et al. “Nivolumab with or without ipilimumab in patients with recurrent glioblastoma: results from exploratory phase I cohorts of CheckMate 143.” Neuro-Oncology, Volume 20, Issue 5, April 2018, Pages 674–686. DOI: 10.1093/neuonc/nox208. Published 2018-04. Accessed 2026-05-09.

doi.org

This paper reports limited clinical activity for PD-1-pathway blockade in recurrent glioblastoma. It is a peer-reviewed journal article, not a preprint. It supports the claim that checkpoint biology does not automatically translate from extracranial melanoma to intracranial glioma; blood-brain barrier penetration and immunosuppressive CNS microenvironment are mechanistic concerns. I do not assert a randomized overall survival benefit here because this source is the exploratory phase I cohort paper, not the separate randomized CheckMate 143 cohort.

4. FDA regulatory record: Keytruda BLA 125514

The FDA regulatory record for pembrolizumab, BLA 125514, establishes the approved product and its original accelerated approval in advanced melanoma on 2014-09-04:

FDA Drugs@FDA, Keytruda BLA 125514. Accessed 2026-05-09.

accessdata.fda.gov

This is a Tier 1 regulator source. It does not by itself establish current indications or performance; labels must be retrieved separately.

5. Company filing cross-reference: Merck 10-K

Merck & Co., Inc.’s Form 10-K for fiscal year 2023 is the primary company filing used to cross-reference marketing narratives against filed risk disclosures:

Merck & Co., Inc. Form 10-K for the fiscal year ended December 31, 2023, filed 2024-02-26, SEC accession 0000310158-24-000014. Accessed 2026-05-09.

sec.gov

The 10-K reports Keytruda 2023 global sales of $25.0 billion and describes Keytruda as a key growth driver. It also contains risk-factor language on patent expirations, competitor products, and payer/government price pressure. This is a Tier 2 company filing — self-reported but primary for commercial/risk statements.

Cross-file contradictions and source-bound claims

1. Clinical efficacy vs. filing risk language. The NEJM paper and FDA approval establish a clinically meaningful survival benefit in melanoma. The Merck 10-K, by contrast, frames Keytruda’s commercial outlook through competitive, pricing, regulatory, and patent-expiry risk factors. There is no numeric contradiction — both can be true — but the filing’s language is more conservative than product marketing. Source-bound claim: efficacy is established in advanced melanoma, but future revenue durability is not guaranteed by the cited 10-K.

1. Patent grant vs. performance. US 8,952,139 B2 is a granted patent, but a grant is a record of allowed claims, not proof of clinical benefit, commercial viability, or freedom to operate. This note does not use the patent grant as performance evidence.

1. CNS extrapolation gap. The CheckMate 143 exploratory paper shows limited activity for PD-1 blockade in recurrent glioblastoma. Any claim that checkpoint inhibition is a proven CNS tumor therapy would be contradicted by this source. The note therefore treats CNS checkpoint therapy as unproven outside specific histologies/tissue-agnostic approvals.

Quantitative worked calculation: number needed to treat from KEYNOTE-006

Claim being tested: How many patients need to be treated with pembrolizumab rather than ipilimumab to have one additional patient alive at 12 months?

Formula:

NNT = 1 / (P_pembro_12mo − P_ipi_12mo)

Inputs, from NEJM DOI 10.1056/NEJMoa1503093 (2015-06-25):

• P_pembro_12mo = 0.741
• P_ipi_12mo = 0.582

Calculation:

Difference = 0.741 − 0.582 = 0.159

NNT = 1 / 0.159 = 6.29

Interpretation: Approximately 6.3 patients treated with pembrolizumab every 2 weeks instead of ipilimumab for one additional patient to be alive at the 12-month landmark. This is a simplified, landmark-based NNT — it is not a formal survival-model NNT, and it does not account for censoring, duration of response, later survival divergence, or quality of life.

Sensitivity:

• If the 12-month OS difference were 0.10, NNT = 10.
• If the 12-month OS difference were 0.20, NNT = 5.

A 5-percentage-point change in the assumed absolute OS benefit changes the NNT by roughly 1.6 to 3.7 patients in this range, so the estimate is moderately sensitive to the exact Kaplan-Meier readings.

Parameter table

Unit economics / business model

Who monetizes: pharmaceutical innovator companies monetize checkpoint inhibitors by selling high-priced biologic drugs to hospitals, clinics, and specialty pharmacies. Merck reported Keytruda sales of $25.0B in 2023 in its 10-K. Product-level gross margins are not separately disclosed in the cited 10-K, so this note does not assert a product-level margin figure. Value capture during patent life is supported by the granted patent record and by Merck’s 10-K risk language on loss of exclusivity; after exclusivity, biosimilar entrants can capture value.

Typical margins: needs primary source validation — the cited Merck 10-K does not break out Keytruda gross margin. Do not use sector-wide gross-margin estimates from secondary sources here.

Who captures value: the patentee/innovator captures value during the exclusivity period; payers and hospitals capture value through reimbursement management; after exclusivity, biosimilar manufacturers and payers capture value. Each of these claims is source-bound to the patent, FDA record, or 10-K where applicable.

Disconfirming Evidence

The thesis — that anti-PD-1 checkpoint blockade is a durable, high-value oncology technology and a plausible CNS tumor strategy — would be falsified or weakened by:

1. Patent invalidation or non-coverage: If US 8,952,139 B2, or the actual Keytruda patent estate, is invalidated or held not to cover pembrolizumab in a PTAB/district-court proceeding, the intellectual-property moat weakens.
2. Negative randomized CNS data: If a randomized CheckMate 143 or equivalent trial confirms no overall survival benefit in glioblastoma, the CNS extension would be disconfirmed. The exploratory cohort already shows limited activity.
3. Reimbursement collapse: If Medicare/Medicaid or private payers impose severe price controls or restrict coverage, the $25B Keytruda revenue base would not repeat. The Merck 10-K risk factors identify this.
4. Faster biosimilar entry: If biosimilar competitors launch earlier than expected after exclusivity, the value-capture model weakens.
5. Superior competing mechanism: If a different mechanism — e.g., neoantigen vaccines, CAR-T, or brain-penetrant small molecules — demonstrates better CNS or checkpoint-resistant tumor efficacy, the centrality of anti-PD-1 in the oncology/CNS portfolio declines.

Function

Seed technology note — created during the 2026-07-27 vault gap audit. This note covers the oncology and cns therapeutics technology space within the Healthspan universe.

Why it's a bottleneck

The moat dynamics, regulatory barriers, and installed-base effects that make this space hard to replicate or substitute — to be researched in deep-dive phase.

Companies

• GILD (Gilead Sciences) — HIV (Biktarvy), oncology (Trodelvy, CAR-T: Yescarta/Tecartus), liver disease
• VRTX (Vertex Pharmaceuticals) — Cystic fibrosis (Trikafta), pain (suzetrigine/NaV1.8), T1D cell therapy (VX-880)
• TGTX (TG Therapeutics) — Anti-CD20 B-cell therapies (BRIUMVI for MS)
• IONS (Ionis Pharmaceuticals) — Antisense RNA therapeutics (TRYNGOLZA, SPINRAZA, WAINUA)
• ARWR (Arrowhead Pharmaceuticals) — RNAi therapeutics (plozasiran for FCS/CVD, zodasiran for lipid disorders)
• BCRX (BioCryst Pharmaceuticals) — Small-molecule rare disease (Orladeyo for HAE)
• ACAD (ACADIA Pharmaceuticals) — NUPLAZID for Parkinson's psychosis, DAYBUE for Rett syndrome

Open questions

• [ ] Which specific SEC filing confirms the revenue breakdown and competitive position for each company in this space?
• [ ] What are the key regulatory moats (FDA approval pathways, CLIA certification, CMS reimbursement) for this category?
• [ ] Which companies have the strongest patent/IP position — what do their patent filings reveal about the competitive landscape?
• [ ] What are the customer switching costs — what makes it hard for buyers to switch to an alternative?

Who buys it

Customer classes named from evidence:

• Hospitals and oncology clinics — the FDA label describes intravenous infusion in healthcare settings; Merck’s 10-K describes sales to wholesalers, hospitals, clinics, and government agencies.
• Patients with advanced melanoma and other approved indications — KEYNOTE-006 enrolled advanced melanoma patients; FDA BLA 125514 is the approval record.
• Government and private payers — Merck’s 10-K risk factors reference Medicare, Medicaid, and private payer price pressure.
• Prescribing oncologists — the NEJM paper reports oncologist-led trial treatment decisions.

No buyer is invented: these classes are drawn from the FDA label, the NEJM paper, and the Merck 10-K.