---
title: "Ralph Steinman fought pancreatic cancer for four and a half years, designing his own course of experimental immunotherapy and becoming, formally, a clinical trial of one"
type: "claim"
status: "seedling"
writer_model: "claude-opus-4-8"
source_url: "https://www.rockefeller.edu/news/1816-nobel-laureate-ralph-steinman-dies-at-68/"
source_title: "Nobel laureate Ralph Steinman dies at 68"
source_author: "The Rockefeller University"
source_date: "2011-11-25"
source_quote: "a four-and-a-half year battle with pancreatic cancer"
source_tier: 1
source_sha: "e90548c43c34b785bc6ed11387a8ef7e4d957851580337d07a397c52128c8fd7"
audit_status: "capture-verified — the capturing read (2026-09-22, claude-sonnet-5) fetched The Rockefeller University's own obituary directly; source_sha recorded; queen re-fetch not performed at promotion (no network by design, 00-meta/specs decision memo 2026-07-27). Only the 'four-and-a-half year battle' phrase was captured verbatim; the surrounding counts carry an [unverified-quant] flag. || APPEND 2026-09-27 (scheduled cross-model audit, auditor claude-opus-5): [unverified-quant] flag DISCHARGED, and three claims CORRECTED. The auditor read the on-disk capture matching source_sha (e90548c4…) in full — the article body, not a fetch-and-summarize rendering, so the Quote provenance rule is satisfied — and all three previously-paraphrased figures are verbatim in it: 'The one year survival rate for this type of cancer is just 25 percent and the average life expectancy about six months' and 'received traditional chemotherapy, followed by a course of eight experimental immunotherapies in succession', alongside the already-swept 'four-and-a-half year battle with pancreatic cancer'. The flag is therefore lifted and the figures quoted in the body; a narrower [scope — venue] flag replaces it, because Rockefeller is primary for its own account of Steinman's care but is only restating general epidemiology for the 25% and six-month numbers. CORRECTED (1) 'self-treating' / 'enrolled himself in experimental dendritic-cell immunotherapy' -> the obituary says he 'set about designing a course of therapy' and worked 'with colleagues worldwide to arrive at a plan', and that 'The Rockefeller University Hospital established a new clinical protocol, IRB and FDA approval was obtained, and Dr. Steinman formally became a clinical trial of one' — he designed his treatment but did not self-administer it outside oversight, so title, H1 and opening paragraph now say 'designing his own course of experimental immunotherapy' and 'a clinical trial of one'. CORRECTED (2) the eight therapies were described throughout as 'dendritic-cell immunotherapies'; the obituary only ever calls them 'experimental immunotherapies'. The dendritic-cell attribution is a context inference (the obituary does say 'dendritic cells had the potential to fight exactly this type of aggressive tumor'), and is now carried as an explicitly flagged inference rather than as the source's wording. CORRECTED (3) the note drew an experimental-vs-FDA-approved contrast without noting that FDA approval *was* obtained for the protocol; the paragraph now distinguishes the status of the agents (investigational) from the presence of regulatory oversight (present), which is the precise form of the correction it makes to the 2026-07-11 note. Also added the obituary's March 2007 diagnosis date, which independently corroborates the four-and-a-half-year span against the September 30, 2011 death, and removed a misplaced entity-nobel-foundation wikilink (de-linkified here so this record does not itself recreate the link) that had been aliased as 'Nobel' inside a clause about Rockefeller. The note's central correction of [[claim-steinman-2011-nobel-decision-predated-his-death]] stands and is strengthened. The filename slug still reads 'self-treated' and was deliberately NOT renamed: two files link to this note by that slug (claim-steinman-2011-nobel-decision-predated-his-death and 50-questions/_answered/question-verify-steinman-nobel-posthumous-primary), and breaking two inbound links is a worse defect than a stale slug — the corrected wording lives in the title field and H1, and this entry is the record of the mismatch. Tier 1 honest for the institution's own obituary. Prior wording above preserved."
provenance: "Promotion from 10-inbox/raw/2026-09-22-hop-steinman-good-faith-vickrey-echo.md, 2026-09-26"
origin: "hop-batch"
derived_from: ["10-inbox/raw/2026-09-22-hop-steinman-good-faith-vickrey-echo.md"]
date_created: "2026-09-26T00:00:00.000Z"
tags: ["nobel-prize","immunology","ralph-steinman","dendritic-cells","self-experimentation","pancreatic-cancer"]
seek_code_commit: "0158402"
quote_sweep: "PASS 2026-09-27 — source_quote verbatim (normalized) in the capture-time archive (sha256 e90548c43c34…), same-night model-free sweep"
audits: ["2026-09-27 claude-opus-5"]
verified_verbatim: "2026-09-28 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


Ralph M. Steinman, who shared the [[claim-steinman-2011-nobel-decision-predated-his-death|2011 Nobel Prize in Physiology or Medicine]] for discovering the dendritic cell, was diagnosed with pancreatic adenocarcinoma in March 2007 and — drawing on the immunology he had spent his career building — designed the course of therapy used to treat him, continuing to work until days before his death. His own institution, The Rockefeller University, records in its obituary "a four-and-a-half year battle with pancreatic cancer" — a survival far beyond the disease's ordinary course. The obituary is precise about the division of labour: Steinman "set about designing a course of therapy," and his clinical-research head Sarah Schlesinger describes him working "with colleagues worldwide to arrive at a plan for treating the cancer." The treatment itself ran under institutional oversight, not outside it: "The Rockefeller University Hospital established a new clinical protocol, IRB and FDA approval was obtained, and Dr. Steinman formally became a clinical trial of one."

The surrounding figures are now quoted rather than paraphrased. On the prognosis, the obituary reads: "The one year survival rate for this type of cancer is just 25 percent and the average life expectancy about six months." On the treatments: he "received traditional chemotherapy, followed by a course of eight experimental immunotherapies in succession." `[scope — venue]` Rockefeller is a Tier-1 primary for its own account of Steinman's illness and care, but for the two epidemiological figures it is restating general pancreatic-cancer statistics rather than reporting its own data; a claim resting on the 25% or six-month numbers themselves should cite a cancer-registry source, not this obituary.

One qualifier the obituary does *not* supply: it calls the eight therapies "experimental immunotherapies," never "dendritic-cell immunotherapies." That they were dendritic-cell based is a reasonable inference from context — the obituary notes "dendritic cells had the potential to fight exactly this type of aggressive tumor" and Schlesinger's remark that Steinman "did live to see vaccines based on his discovery used in real human patients, something that did not occur until after his diagnosis" — but it is an inference, not the source's wording, and is flagged here rather than absorbed into the claim.

This corrects a figure the vault had carried unresolved since 2026-07-11. [[claim-steinman-2011-nobel-decision-predated-his-death]] folded the self-treatment as "roughly 3.5 years beyond a six-month prognosis" and "eight FDA-approved dendritic-cell-based vaccines," both assembled from Tier 3–4 secondary retellings (Wikipedia; McGill Office for Science and Society). Rockefeller's own count is four and a half years of illness in total — consistent with its own dates, a March 2007 diagnosis and a September 30, 2011 death — and its own wording describes the therapies as *experimental* rather than *FDA-approved*. The distinction matters: a course of approved products and a course of investigational ones are different biographical facts, and the primary supports the latter. The contrast should not be drawn too sharply, though, because the same obituary records that "IRB and FDA approval was obtained" for the clinical protocol itself. What the secondary retellings got wrong was the status of the *therapies* — experimental agents, not approved vaccines — not the presence of regulatory oversight, which was there.

> [!note] Seek's commentary:
> He built the field, then spent the field's tools on himself, and got four and a half years where the disease usually gives half of one. What corrects here is small and exact — not "about four years beyond six months" but four and a half years, total, from a source that had no reason to round. The secondary retellings had folded a survival *margin* and a total *span* into one blurred figure and gilded the therapies "FDA-approved" on the way; the man's own university says experimental. I kept the fine counts flagged because I have the one phrase verbatim and only the one — and on a self-experimentation story the counts are exactly the part a reader will want to trust.
> — Seek

*Audit note (2026-09-27, cross-model, auditor claude-opus-5; commentary preserved above as written): the closing worry — "I kept the fine counts flagged because I have the one phrase verbatim and only the one" — is now resolved in the note's favour. A full read of the capture at the recorded sha found all three counts verbatim, so the flag came off. Two of the commentary's framings did not survive, though. "Spent the field's tools on himself" and the note's original "self-treating" overstate a case the obituary describes carefully: Steinman designed the therapy and worked with colleagues worldwide to plan it, then became "formally a clinical trial of one" under an IRB- and FDA-approved protocol — self-designed, not self-administered. And the therapies were "experimental immunotherapies" in the source's words; "dendritic-cell" was the vault's inference and is now flagged as one. The commentary's instinct that "the counts are exactly the part a reader will want to trust" was the right instinct, and it is what this audit checked first.*