19.1 — What Cancer Is

Around half of people in high-income countries will be diagnosed with cancer at some point.

And around half of those diagnosed will not die of it. Cancer survival in the UK has roughly doubled over the last fifty years, and for several cancers the change has been far larger than that.

Cancer is not one disease. It is a category containing over 200 conditions with different causes, behaviours and outlooks — which is why a single word covering both a basal cell carcinoma cured by a five-minute excision and a pancreatic cancer is so misleading.

What it actually is

Cancer is a failure of the controls that decide whether a cell divides (Chapter 1.7).

And it is an evolutionary process, running inside your body.

A population of cells with heritable variation, competing for resources, under selection. Cells with mutations that help them survive and divide outcompete their neighbours. The tumour that emerges is the product of natural selection operating over years (Chapter 3.2).

Which explains everything that makes cancer difficult: why it becomes more aggressive over time, why it develops resistance to treatment, and why it varies so much between patients and even within a single tumour.

The controls that fail:

Proto-oncogenes — the accelerator. One faulty copy is enough, because a stuck accelerator does not care that the other pedal works.

Tumour suppressor genes — the brakes. Both copies must fail (Chapter 1.7).

DNA repair genes — and their failure produces a much higher mutation rate, which accelerates everything else (Chapter 2.2).

And it takes multiple hits. A typical solid tumour requires 5 to 8 driver mutations accumulated over years to decades. Which is why cancer is predominantly a disease of ageing — time is what allows mutations to accumulate.

The hallmarks

Douglas Hanahan and Robert Weinberg set out a framework in 2000, updated since, describing the capabilities a cell must acquire.

Sustained growth signalling — dividing without being told to.

Insensitivity to anti-growth signals — ignoring the brakes.

Evading apoptosis — refusing the instruction to die (Chapter 1.8).

Limitless replication — reactivating telomerase, which around 85 to 90 percent of cancers do (Chapter 1.7).

Angiogenesis — recruiting a blood supply, without which a tumour cannot grow beyond about 1 to 2 millimetres.

Invasion and metastasis — Chapter 19.3.

And four added later:

Altered metabolism — the Warburg effect: tumours use glycolysis even when oxygen is plentiful, which is inefficient and is apparently useful for generating building blocks. It is also what makes PET scanning work (Chapter 16.6).

Evading immune destruction — Chapter 19.7.

Genome instability — the enabling characteristic that generates the rest.

Tumour-promoting inflammation (Chapter 16.3).

What causes it

And the honest answer includes a substantial component of chance, which people find harder to accept than either genes or lifestyle.

Random replication errors. Every cell division carries a small chance of a mutation (Chapter 2.2), and tissues that divide more accumulate more.

A 2015 analysis found that variation in cancer risk between tissues correlated strongly with the number of stem cell divisions in that tissue — which was widely reported as "most cancer is bad luck". The interpretation was contested, and the honest position is that chance, heredity and environment all contribute, in proportions that differ by cancer type.

And the practical implication matters: a person who develops cancer with no risk factors has not done anything wrong. The tendency to search for a cause, and to find one in the patient's behaviour, is both common and frequently unfounded.

Inherited susceptibility — around 5 to 10 percent.

BRCA1 and BRCA2 — breast, ovarian, prostate, pancreatic. Lynch syndrome — bowel, endometrial (Chapter 2.2). Familial adenomatous polyposis. Li-Fraumeni (Chapter 1.7).

And inheriting a faulty tumour suppressor means being born with one brake already broken in every cell — so only one further hit is needed anywhere, rather than two in the same cell (Chapter 1.7).

Environmental and lifestyle — around 40 percent of cancers are attributable to modifiable factors.

Tobacco — the largest single cause. Responsible for around 20 to 25 percent of all cancer deaths, and not only lung: mouth, throat, oesophagus, stomach, pancreas, bladder, kidney and cervix.

Obesity — at least 13 cancers (Chapter 18.8).

Alcohol — mouth, throat, oesophagus, liver, breast and bowel. Classified as a Group 1 carcinogen, and the risk begins at low intakes for breast cancer.

Infection — around 13 percent of cancers worldwide. HPV, hepatitis B and C, H. pylori, Epstein–Barr virus, HIV indirectly, and schistosomiasis.

And this is where prevention is most direct. Vaccination against HPV and hepatitis B prevents cancer (Chapters 13.5, 17.10), and eradicating H. pylori reduces gastric cancer (Chapter 9.2).

Ultraviolet radiation — skin cancer (Chapter 14.5).

Ionising radiation, air pollution, and specific occupational exposures — asbestos, benzene, aromatic amines.

Diet — processed meat is classified as a Group 1 carcinogen for bowel cancer, and red meat as Group 2A.

And the classification is frequently misread. Group 1 means the evidence that it causes cancer is strong — not that the effect is large. Processed meat and tobacco are both Group 1, and tobacco raises lung cancer risk twentyfold while processed meat raises bowel cancer risk by around 18 percent per 50 g daily. The category describes certainty, not magnitude.

What does not cause cancer

Worth stating, because misinformation in this area is abundant and causes real anxiety.

Mobile phones and Wi-Fi. Non-ionising radiation does not have enough energy to damage DNA, and large epidemiological studies have found no association. Brain tumour incidence has not risen with mobile phone use.

Power lines.

Microwaves, and food cooked in them.

Deodorants and antiperspirants, and underwired bras.

Sugar "feeding" cancer in the sense implied. All cells use glucose. Tumours use more, which is why PET works — and cutting dietary sugar does not starve a tumour, because the body maintains blood glucose regardless.

Artificial sweeteners — at normal intakes, no established human cancer risk, and the aspartame classification in 2023 was as "possibly carcinogenic" on limited evidence, with the acceptable daily intake unchanged.

Acidic diets and "alkalising" — blood pH is tightly regulated and unaffected by diet (Chapter 10.4).

Stress causing cancer directly — the evidence does not support it, though stress affects behaviours that do.

Benign and malignant

Benign — grows locally, does not invade, does not spread, usually encapsulated, and slow.

Malignant — invades, spreads, poorly circumscribed, and grows faster.

And benign is not the same as harmless. A benign brain tumour in a confined space can be fatal, and a benign tumour producing a hormone can be dangerous.

The naming, which is otherwise baffling:

-oma — usually benign. Lipoma, adenoma, fibroma. Carcinoma — malignant, from epithelium. 80 to 90 percent of cancers (Chapter 4.2). Sarcoma — malignant, from connective tissue. Rare, younger patients, spreads through blood rather than lymph. Leukaemia — blood-forming tissue. Lymphoma — lymphoid tissue. Melanoma — melanocytes.

And the exceptions that break the rule: melanoma, lymphoma and glioma are all malignant despite the -oma ending, which is a historical accident rather than a system.

The encouraging framing

Because this Part should not read as a catalogue of threats.

Around 40 percent of cancers are preventable — through not smoking, weight, alcohol, sun protection, vaccination and screening.

Screening prevents cancers rather than merely finding them. Removing a colonic polyp prevents the cancer it would have become (Chapter 9.6). HPV vaccination is on course to eliminate cervical cancer (Chapter 15.13).

Survival has improved substantially. Testicular cancer: over 95 percent cured, including most metastatic disease. Childhood leukaemia: from under 10 percent survival in the 1960s to over 90 percent. Breast cancer: five-year survival above 85 percent in most high-income countries. Melanoma caught early: over 95 percent.

And metastatic melanoma, which had a median survival under a year a decade ago, now has long-term survivors (Chapter 19.7).

Cancer remains serious. It is also, increasingly, survivable — and for a growing number of cancers, curable.

What the next page fixes

Chapter 19.2 covers the hallmarks in more detail — the specific capabilities a normal cell must acquire, and how each of them has become a treatment target.