A working reference
Breast Cancer
Biology, heterogeneity, and the decisions that follow. Written for people who already treat this disease, and arranged so that the biology is carried once, in full, and the management chapters cite it rather than restate it.
A breast cancer is never one thing, and the report in front of you is a sample of it. Most of the hard decisions in this book follow from taking that seriously.
The figure: four clonal populations in overlapping territories, and a core passing over them. What the core holds is what the report would say. Move it and the answer moves, which is the argument the sixth part of this book makes at length.
How it is arranged
Two decisions shape every chapter, and they are worth knowing before you start.
Mechanism lives in one place
Drug class, modality and pathway are written out in full in Part VIII. The disease chapters cite that material instead of repeating it, which is what lets a decision chapter stay short enough to read between clinics. Trial results are not written out at all: they are generated from one registry, so a new readout is entered once and appears in every chapter it bears on.
The assay is not the tumor
Every chapter separates what is biology from what is measurement. A threshold, a sampling scheme and a scoring convention are decisions someone made, and most of the disagreement in this field lives in that gap rather than in the tissue.
The parts
In reading order. History and biology first, then measurement, then treatment, then everything that happens to a patient around it. The chapter count beside each part is computed when the book is built.
Two centuries of being confidently wrong about where breast cancer goes and why. Each idea is read for the error it corrected and the one it introduced.
What the normal breast does, and how much of it a tumor keeps. Precursor states come before any malignant biology, because that is the order the tissue works in.
Every taxonomy in this field was built for a purpose. The useful question is not whether a classification is true but what it was built to decide.
Organized by pathway rather than by drug. Each chapter closes on the resistance mechanisms that the metastatic chapters will have to act on.
Compartments first, then function, then spatial arrangement, then the host. Checkpoints arrive late here on purpose.
The core of the book. It sits before the treatment chapters so that all of them can assume it, and every chapter in it separates what is biology from what is measurement.
Every assay here is a sampling procedure. What it measures, what it stands in for, and how far apart those two can drift.
Modality and drug class, carried once and in full. The disease chapters cite this part instead of repeating it, which is what keeps them short.
Setting-level chapters. What to do about a lesion that may not need treating, about disease that presents late, about a blood test that finds disease no scan can, and about the neoadjuvant window as an experiment rather than a schedule.
What holds across every subtype once the disease is incurable. Goals, burden, and the organ-specific problems that do not sort by receptor.
The commonest subtype, read end to end rather than twice. Endocrine dependence is the thread the whole part pulls on: what establishes it, what sustains it, how it fails, and what is left when it has. The synopsis states the spine; the chapters argue it.
One target and two decades of drugs against it. This is the subtype whose neoadjuvant platform taught the rest of the field how to de-escalate, and the only one where a second receptor changes the question rather than the dose. Triple-positive disease sits here.
A subtype named for what it lacks, which is why its treatment history is the history of finding something to aim at. Chemotherapy, then the immune microenvironment, then the antigens a conjugate can reach.
The patients trials enroll least and clinics see often enough. Extrapolation is the default here, so each chapter says how far it is being stretched.
A full part rather than an appendix. Most of what a patient actually lives with sits here.
Biology and access are separated at every step, and race is not allowed to stand in for either.
The part that will age fastest. Organized by mechanism rather than by agent, so that a new drug does not require a new chapter.
How it is built
Every part, chapter and section carries a permanent identifier that says nothing about where it sits. A chapter keeps its identifier when it moves, and its file is named for it, so cross-references and citations survive any amount of reordering. Part numbers, chapter numbers and section numbers are computed when the book is built and exist nowhere in the source.
Citations go through a single reference store. A chapter names a study by a permanent key and the printed number is worked out per page, so moving a chapter changes the numeral beside a citation and nothing else.
The whole book installs to a home screen and reads with the network switched off. Nothing on any page is fetched from anywhere.