VOLUME I · THE UNSTABLE OBJECT ·
How to Read This Volume
This book begins with a problem of identity. A living medicine is not a stable object that can always be described independently of the route by which it was made, handled, and administered. Its history matters. The argument follows that fact through biological identity, clinical proof, regulatory decisions, publication, payment, consent, and long-term responsibility.
Part One asks what kind of object a cellular treatment is. Part Two examines what that object does to clinical evidence. Part Three considers the institutions that must decide whether a product can be trusted, delivered, paid for, and explained to a patient.
Each of the eighteen chapters preserves the original argument while adding a new section grounded in current public records. The appendices then extend the discussion into product-level teaching files, institutional decisions, source interpretation, patient handoffs, research questions, and a working glossary.
A product name, an approval date, or an impressive response percentage is never enough on its own. Ask which patients were counted, what was measured, where the cells came from, what processing occurred, how the treatment was delivered, and how long anyone actually watched.
Statements about current indications reflect the public record reviewed through August 2026. Earlier approval documents are used for historical questions; current treatment questions must be answered from the current label. Draft agency guidance is identified as guidance, not presented as binding law.
The audit sequence, proposed gates, and persistence categories are the author’s analytical proposals. They are not adopted regulatory instruments and do not replace product-specific clinical judgment, applicable law, or formal institutional procedures.
Illustrations in this volume are original explanations rather than reproductions of manufacturer material or substitutes for primary clinical data. Where a diagram includes patient counts or a dated event, the underlying number is tied to the record discussed in the adjacent chapter.
Read each product file as a bounded study of a particular intervention, population, and public record. A shared disease, cell type, or commercial description does not establish that another preparation has the same biological behavior, evidence, or approved use.
When a regulatory decision includes a younger age group, distinguish direct observation from extrapolation. When a warning changes, separate the original historical decision from the current indication, the event actually observed, and the question of individual causation.
Payment, fertility, treatment travel, manufacturing delay, patient withdrawal, and later clinical surveillance belong to the real treatment pathway. They are not secondary concerns to be considered only after a favorable efficacy percentage has been quoted.
Standing Definitions and Reading Rules
The audit sequence, gate rule, and persistence classes below are proposals advanced by this work. They have no independent regulatory standing and should be tested against the product-specific record instead of treating them as adopted standards.
The Product-Identity Audit Sequence
- Source material: donor or patient, tissue of origin, and collection route.
- Manipulation: editing, transduction, selection, expansion, and cryopreservation history.
- Intended biological action: what the cells are supposed to do and where.
- Release potency: which test supports the decision to administer a lot.
- Delivery route and dose, including the stage at which dose was counted.
- Recipient conditioning and immune management.
- Indication stated at the level of the current approved label or defined study.
- Endpoint and the period during which it was observed.
- Durability distinguished from prediction or biological persistence.
- Adverse-event pathway, including plausible late risks.
- Manufacturing reliability and losses before infusion.
- Cost, access, fertility consequences, and patient burden.
The Gate Rule
A composite assessment is not interpreted until the necessary conditions have been examined. If product identity, meaningful release testing, relevant safety characterization, or persistence-matched follow-up cannot be established or explicitly bounded, a numerical score cannot repair the missing condition.
Three Persistence Classes
Transient. Cells are cleared over days to weeks; the relevant risk may be concentrated near administration, although a biological effect can continue.
Persistent, non-renewing. Cells remain for months or years without evidence of continuing self-renewal; exposure continues while the cells survive.
Self-renewing. A cell population can persist through continuing replication; monitoring obligations must account for a potentially open-ended exposure.
Preface
The living-medicine record changed while this book was being prepared. New safety findings narrowed one product’s appropriate use. A federal payment model began addressing fertility and access directly. In July 2026 the approved age range for another treatment expanded to children as young as two, with the youngest patients reached through stated extrapolation. Those events did not change the central argument. They showed why it matters. Nobody has ever seen the whole medicine.
That sentence opens Chapter One and it is the reason these three volumes exist. A pharmacist handed a vial of atorvastatin can verify the contents. A clinician handed a vial of cellular therapy cannot, because there is no single invariant molecular structure capable of establishing the identity of the entire cellular population — only a population of cells carrying the age of a donor, the injuries of a collection, the pressures of a culture, and the damage of a freeze. From that single fact a great deal follows, and most of it has not been followed.
Medicine’s institutions were built around an assumption so basic it was never written down: that a medicine is a stable object whose identity can be established independently of how it was made. Regulation assumes it. Peer review assumes it. Reimbursement, liability, and everyday clinical inference all assume it. None of them announces the assumption, and none has an alarm that sounds when it fails. Each simply continues operating, producing outputs that look entirely normal and mean something narrower than their readers believe. This series is an attempt to work out what that costs and what should replace it.
Volume One makes the argument. Eighteen chapters on what kind of object a cellular therapy is, what its instability does to the machinery of clinical proof, and how five institutions have responded — regulation best, peer review worst, and payment with the least excuse. It closes by converting the argument into eight conditions under which a claim about a living medicine could be believed.
Volume Two tests those conditions against fourteen indication files, built to an identical shape so they can be compared. Six that hold, four still forming, four that failed. The failed files are the point. A work examining only successes cannot teach anyone to recognize a failure while it is still happening, and the field’s most damaging episode ran for a decade inside a functioning peer review system.
Volume Three states what an institution must build: the chain of identity, potency assurance, the release decision, the accounting for failure before infusion, the decision instrument with its gates and its stated failure conditions, surveillance as infrastructure instead of an obligation, and finally the conditions under which a center has earned the right to treat anyone at all. Three things should be said plainly about the standing of what follows.
This book supports the responsible development and use of these therapies. At FDA approval, twenty-nine of thirty-one efficacy-evaluable patients achieved freedom from severe vaso-occlusive crises for at least twelve consecutive months. Thirty patients constituted the prespecified primary efficacy population; FDA included one additional patient as a nonresponder in the approval analysis. That result, against a pretreatment burden of repeated severe crises, represents a substantial clinical effect. The argument of this volume is for stronger evidence and more complete disclosure, not less therapeutic progress.
Its criticism is directed at the evidentiary and institutional translations surrounding these treatments rather than at the treatments themselves or at the good faith of regulators. On the evidence assembled in Part Three, the regulator has read this problem more accurately than the journals, the payers, or the field’s own literature, and arrived at a persistence-based follow-up taxonomy before anyone else did.
Every claim in this volume remains subordinate to the public record. Claims were subjected to a defined verification procedure, and where verification narrowed or contradicted the draft, the record controlled.
What the argument requires is not less confidence in cellular therapy but a different basis for it. Confidence in a molecule can rest on the object. Confidence in a living medicine has to rest on the record: where the cells came from, what was done to them, what was measured, what was not measured, who was lost between consent and infusion, and how long anyone has actually been watching.
So the title’s claim is a literal one. Cells remember, and no current release test can reconstruct the complete biological and manufacturing history they carry. The trustworthiness of a cellular therapy is therefore a property of the documentation surrounding it instead of the vial. An unusual claim. The three volumes are an attempt to make it stick.
Source note: Direct serialization of the book’s “How to Read This Volume,” “Standing Definitions and Reading Rules,” and “Preface.”
