The American Cancer Society estimates that roughly half of all U.S. cancer deaths are attributable to modifiable risk factors. The W-H-O estimates 30 to 50 percent of the global cancer burden is preventable with tools we already have. This chapter is about where in the disease process the intervention is cheapest and most effective — and about why we so consistently underinvest there.

Public health sorts prevention by what stage of disease the intervention targets, and the level determines what is acceptable to do. Primary prevention stops disease from ever occurring: not smoking, avoiding oncogenic infections, limiting carcinogen exposure. The target is a healthy person — and that target sets a severe safety requirement. You are intervening on millions who would never have gotten cancer anyway. Secondary prevention detects disease at an early, more curable stage. Tertiary prevention manages established disease. The recurring policy error is confusing the levels.

If you wanted to save the most lives from cancer with a single policy lever, the answer is tobacco control — and has been for seventy years. Tobacco causes about 22 percent of cancer deaths globally. In the U.S., lung cancer kills about 140,000 people a year, roughly 90 percent tobacco-related. Eliminating tobacco would prevent more annual cancer deaths than every cancer therapy ever developed, combined. That is not rhetoric. It is arithmetic. Prevention operates at three points: preventing initiation in young people, promoting cessation in current users, and harm reduction for those who will not quit.

About 13 percent of cancer globally is caused by infection, and most of that fraction is preventable by vaccination or by eliminating the infection. The H-P-V vaccine targets the major oncogenic types responsible for most cervical cancer and substantial fractions of anal, oropharyngeal, and other cancers. In countries with high coverage, vaccine-type H-P-V prevalence in young women has fallen 80 to 90 percent. Hepatitis B vaccine, available since 1982, prevents most H-B-V-driven liver cancer — 700,000 deaths a year globally. Hepatitis C is now curable; H. pylori eradication prevents gastric cancer. The tools are inexpensive. The barriers are organizational and political.

Chemoprevention is the use of drugs to prevent cancer in people at elevated risk — the part of primary prevention that most resembles clinical medicine. The evidence for selected agents is strong. Aspirin reduces colorectal cancer incidence and mortality by roughly 20 to 30 percent. Tamoxifen and raloxifene cut invasive breast cancer incidence by about half over five years in high-risk women. Aromatase inhibitors are more effective still in postmenopausal high-risk women. The hard constraint in each case: side effects that are acceptable in someone facing serious cancer risk are not acceptable in someone who would probably never have gotten cancer. That constraint produces the defining discipline of chemoprevention — risk stratification.

For people with hereditary cancer syndromes, primary prevention operates in a more aggressive register, because the cancer risk is not statistical — it is near-certain over a lifetime. BRCA1 and 2 carriers face lifetime breast cancer risks of 65 to 80 percent and ovarian risks of 40 to 60 percent. Options span surveillance, chemoprevention, and prophylactic surgery — bilateral mastectomy cutting breast cancer risk by over 90 percent. Lynch syndrome carriers face elevated risks of colorectal, endometrial, ovarian, and gastric cancers, managed with intensive colonoscopy and in some cases aspirin. Familial adenomatous polyposis requires prophylactic colectomy — the colon must come out.

The pattern of underinvestment in prevention is not mysterious once you understand how health systems pay for things. A dramatic late intervention — surgery, chemotherapy, a targeted drug — is billed at the moment it is delivered, attributed to a specific patient, and visible as a saved life. A quiet intervention — a tax, a vaccine, a smoke-free law — prevents a disease that was going to happen to an anonymous person twenty years from now. The prevented case never appears on any ledger. The health system that fails to prevent it still gets paid when it treats it. The biology is settled. The gap is political economy.

So here is the chapter's claim. Primary prevention — especially tobacco control and oncogenic-virus vaccination — is the highest-leverage way to reduce cancer mortality, far exceeding what treatment achieves per death prevented, and it is underused for reasons of political economy rather than weak evidence. The finding that would force a revision: rigorous long-horizon data showing that high-coverage H-P-V and hepatitis B vaccination programs failed to reduce their target cancers as vaccinated cohorts aged. The early Australian and hepatitis B data point the other way. But the decisive cohorts are still aging — we are running experiments whose results our successors will read.