The Emperor of All Maladies

Mukherjee's conceit is that cancer has a life story — an identity, a character, a history of its relationships with the people trying to kill it — and the book is structured accordingly.

The device works because cancer genuinely changed in what it meant across the period covered: from an unmentionable and uniformly fatal diagnosis to a set of diseases with different mechanisms and very different prognoses.

Cancer changed in meaning across the period, which is what makes a biography the right form.

Devmark’s reading — not a quotation from the book

His opening narrative: a pathologist who had not treated living patients attempting chemotherapy on children with leukaemia, achieving temporary remissions, and watching every one of them relapse.

Those first remissions are the book's founding moment — proof that a chemical could push back a cancer, in a context where nothing could — and the relapses establish the pattern the rest of the book follows.

His account of the radical mastectomy: the theory that cancer spreads centrifugally and that cure requires removing enough tissue, and the increasingly extensive operations that followed from it.

His treatment of how the theory persisted for decades without a controlled trial, and of the surgeons who resisted testing it, is the book's clearest case study in how a plausible mechanism can outrun its evidence and cause enormous harm.

His account of the political campaign that produced federal funding at an unprecedented scale, driven by a lobbyist rather than by scientists.

His assessment is mixed and honest. The funding was transformative; the promise that a cure would follow from resources alone was mistaken, because the underlying biology was not understood well enough for a directed assault to work.

His account of the insight that changed leukaemia outcomes: multiple drugs with different mechanisms given together, at doses that took patients to the edge of death, to prevent resistance from emerging.

His description of what the early protocols cost the children who received them is unflinching, and he does not let the eventual success sanitise the process by which it was reached.

His chapters on the epidemiological work linking smoking to lung cancer, and on the decades of industry obstruction that followed.

His framing of prevention as the intervention that saved the most lives, while receiving a fraction of the attention and funding given to treatment, is one of the book's strongest arguments and applies well beyond cancer.

His account of the convergence: viral research, chemical carcinogens and inherited syndromes all pointing toward mutations in genes controlling cell division.

The discovery that cancer is a disease of the genome — that the causes identified separately were converging on the same machinery — is the intellectual climax of the book and reframes everything preceding it.

His final movement: imatinib for chronic myeloid leukaemia and trastuzumab for a subset of breast cancers, drugs designed against a specific molecular lesion rather than against dividing cells generally.

His honesty about the limits is what distinguishes this section. Targeted drugs work spectacularly for some cancers, resistance emerges, and the approach has not generalised as quickly as the first successes suggested.

The first targeted drugs worked spectacularly. That they did not generalise quickly is the part usually omitted.

Devmark’s reading — not a quotation from the book

Throughout, he returns to individuals under his own care, including one patient whose course runs across the whole book.

This is what makes it readable rather than a history of protocols. It also imposes a discipline on the narrative: every scientific development is assessed by what it meant for someone receiving it.

It is long, and its focus is heavily American — the institutions, the funding battles and the regulatory context are those of one country.

Its final chapters predate the immunotherapy results that have since changed outcomes for several cancers, so a reader should treat the closing assessment of where treatment stands as a snapshot of its moment rather than a current picture.