Cancer epidemiology, screening, and prevention

US cancer epidemiology (2026 estimates)

• Total: ~2.11 million new cases, ~626,000 deaths/yr (ACS 2026).
• Most common new cases (men): prostate > lung > colorectal. (women): breast > lung > colorectal.
• Most common cancer deaths (both sexes combined): lung > colorectal > pancreatic > breast > prostate. By sex: men lung > prostate; women lung > breast.
• Lifetime risk (US): ~40% men, ~39% women.
• Age is the dominant risk factor: SEER (2013 to 2017) incidence was ~8.5-fold higher and mortality ~19-fold higher in those ≥65 vs <65 yr. ~60% of new cancers and ~70% of cancer deaths occur at age ≥65.

Cancer statistics: core measures

• Population rates are expressed as cases per 100,000 per year and are age-adjusted (to a standard population) so populations with different age structures can be compared.
• Data sources: NCI SEER registries (cover ~48% of the US population); CDC National Program of Cancer Registries (46 states + DC; with SEER covers the entire US); mortality from state vital statistics.

Distribution of cancer (person, place, time)

• Person: overall rates are ~13% higher (incidence) and ~35% higher (mortality) in men than women. By race/sex: Black men have the highest cancer risk; Black women have lower incidence than White women (Asian/Pacific Islander and Hispanic women have the lowest) but the poorest survival (~14% lower than White women), yielding mortality ~12% higher than White women. This defines key disparity populations.
• Place: geographic variation (e.g., highest US incidence/mortality in Kentucky, driven by historically high smoking; low in Utah, low smoking/alcohol prevalence).
• Time: trend analysis generates etiologic hypotheses. Examples: prostate incidence rose/fell with PSA screening; lung cancer mortality rose with cigarette use then fell ~20 yr after the 1964 Surgeon General report (delayed ~2 decades in women); gastric cancer fell (sanitation, then H. pylori decline); cervical mortality fell with Pap testing; breast mortality fell with mammography plus adjuvant therapy.

Determinants of cancer and study designs

• Cause (epidemiologic): a factor that increases disease frequency when present and decreases it when absent. Causation is multifactorial; a "necessary cause" is rare (e.g., oncogenic HPV for cervical cancer).
• Hierarchy of evidence: randomized trials give the strongest causal evidence but are ethical only for potentially beneficial exposures, so most etiologic evidence is observational (cohort, case-control, cross-sectional), balancing time/resources against susceptibility to bias.
• Molecular epidemiology: merges population and laboratory tools; uses biomarkers of exposure, internal dose, biologically effective dose (e.g., DNA adducts), preclinical effect, and early detection. Underlies precision prevention/screening (targeting the highest-risk strata).

Carcinogenesis continuum and host genetic variation

• Steps: initiation (inherited or acquired genetic damage from physical, infectious, or chemical carcinogens) → promotion (e.g., androgens in prostate, estrogen in breast/endometrium) → progression. Complete carcinogens (some cigarette-smoke components) act as both initiators and promoters.
• Genetic susceptibility spectrum: rare high-penetrance mutations (e.g., BRCA1/2), large individual risk; common low-penetrance polymorphisms (GWAS-identified, RR ~1.1 to 1.5), small individual but potentially large population effect (used in polygenic risk scores, mostly validated in European-ancestry populations).

Modifiable risk factors

• Tobacco: causally linked to 13 malignancies (lung, oral cavity, larynx, esophagus, bladder, kidney, pancreas, stomach, cervix, colon, rectum, liver, and AML); an estimated ~30% of US cancer deaths (about 19% of cases) are attributable to cigarette smoking. #1 modifiable cause of cancer death.
• Obesity (BMI ≥30): convincingly linked to up to 13 cancers (endometrial, esophageal adeno, gastric cardia, liver, kidney, pancreatic, colorectal, gallbladder, postmenopausal breast, ovarian, thyroid, plus multiple myeloma and meningioma). Also raises recurrence and cancer-specific mortality. Intentional weight loss and bariatric surgery reduce risk.
• Alcohol: causes oral cavity, pharynx, larynx, esophagus (squamous), colorectal, liver, and female breast cancers; ~4% of incident cancers globally (4.1% in 2020); ~401,000 alcohol-attributable cancer deaths in 2019. Synergistic with tobacco for head/neck cancers (combined RR far exceeds additive). Limit: ≤2 drinks/day (men), ≤1 (women).
• Diet: processed meat convincingly raises colorectal risk (IARC group 1); red meat is probable (IARC group 2A); whole grains and fiber lower it. Most vitamin/supplement trials show no benefit (VITAL: no effect of vitamin D or omega-3 on cancer incidence); some show harm.
• Physical inactivity: ~25% of the US population is sedentary; up to ~5% of cancers. Raises colorectal, postmenopausal breast, and endometrial risk; activity also aids survivors.
• UV / ionizing radiation: UV (sun, tanning beds) causes melanoma and nonmelanoma skin cancers; ionizing radiation causes leukemia, thyroid, and female breast cancers.
• Air pollution: outdoor air pollution and particulate matter classified IARC group 1 carcinogens.
• Occupation: ~2 to 8% of cancers (historically, Pott's scrotal SCC in chimney sweeps).

Infection-associated cancers (~15 to 18% of cancers worldwide)

• HPV: cervical, anal, oropharyngeal, vulvar, vaginal, penile.
• HBV/HCV: hepatocellular carcinoma.
• H. pylori: gastric adenocarcinoma, MALT lymphoma.
• EBV: nasopharyngeal, Burkitt, post-transplant lymphoma, gastric.
• HHV-8: Kaposi sarcoma, primary effusion lymphoma.
• HIV: Kaposi, NHL, cervical (AIDS-defining); increases many others.
• HTLV-1: adult T-cell leukemia/lymphoma.

Screening: rationale and principles

• Definition: detect noninvasive or invasive neoplasia in asymptomatic people to reduce morbidity/mortality. A screening test is not diagnostic; it flags high probability of disease warranting further testing.
• Key criterion: the only justification for a population screening recommendation is a demonstrated reduction in cancer mortality (ideally in a randomized trial). Detecting more localized cancers does NOT by itself prove benefit.
• Convincing benefit exists for cervical, colorectal, and breast cancers in average-risk adults, and for lung cancer only in high-risk adults (heavy smokers, not average risk).

Screening biases and harms

• These biases are minimized by a randomized trial with a mortality endpoint. Other harms: complications of the test, workup of false (and true) positives, and overtreatment, plus financial and emotional costs.

Screening test accuracy

• Sensitivity and specificity are relatively independent of prevalence.
• Positive predictive value (PPV) depends heavily on prevalence: screening is most efficient when targeting a common cancer or a known high-risk subgroup.
• An intrinsically ineffective test does not become effective in a high-risk group (e.g., chest radiography did not reduce lung cancer mortality even in heavy smokers). Conversely, an effective test is preferably applied to higher-risk populations.

Major screening programs (USPSTF current)

• Global-specific programs: oral visual inspection (India, reduces mortality ~1/3 in tobacco/alcohol users); HCC US/AFP or MRI (East Asia/Africa); upper-GI/endoscopy for gastric (Japan) and esophageal SCC (China); plasma EBV DNA for nasopharyngeal carcinoma (endemic regions).

Chemoprevention and vaccines by organ site

• Failed / harmful preventives: beta-carotene increased lung cancer in smokers (ATBC, CARET); selenium and vitamin E did not reduce prostate cancer (SELECT); vitamin E raised prostate cancer risk (~17%) and selenium raised high-grade risk in men with high baseline selenium; vitamin D and omega-3 no benefit (VITAL); high-dose 13-cis-retinoic acid too toxic and ineffective at lower doses.

Vaccination and cancer prevention

• HPV vaccine (Gardasil 9): covers 9 HPV types; ages 9 to 26 (catch-up through 26; ages 27 to 45 by shared clinical decision-making). Prevents ~6 cancers.
• HBV vaccine: prevents HBV, reducing HCC.

Multi-cancer early detection (MCED)

• Galleri (Grail): cell-free DNA methylation analysis, detects signal across >50 cancer types. PATHFINDER (Schrag Lancet 2023): 6,621 asymptomatic adults ≥50; 1.4% positivity, 38% PPV. SYMPLIFY (UK symptomatic): 66% sensitivity, 75% PPV, 98% NPV.
• Status: not yet FDA-approved (PMA submitted Jan 2026; FDA advisory committee voted in favor Sep 2026, decision pending); available as a laboratory-developed test. NHS-Galleri RCT (~142,000): primary endpoint (stage III/IV reduction) not met, stage IV reduced ~14% (ASCO 2026); no mortality data yet.
• Caveats: limited early-stage sensitivity; low-shedding tumors missed; false positives drive costly workups; not a substitute for established screening (mammography, colonoscopy, LDCT, Pap).
• Other MCED: CancerSEEK, DELFI fragmentomics (investigational).

Hereditary cancer overview (see Genetic Testing section)

• BRCA1/2, Lynch, FAP, Li-Fraumeni, Peutz-Jeghers, Cowden, von Hippel-Lindau, NF1/2, MEN1/2: most testable hereditary syndromes.
• Universal tumor MMR/MSI testing on all colorectal and endometrial cancers.
• Universal germline testing for ovarian, pancreatic, and metastatic prostate cancers.

2024-2026 prevention & screening updates

• Self-collected HPV cervical screening: FDA cleared May 2024 (Roche cobas HPV self-collect; BD Onclarity), in-clinic patient-collected vaginal swab expands access.
• HPV single-dose vaccination: WHO 2022 endorsed single-dose 9vHPV for ages 9 to 20 (high efficacy against persistent vaccine-type HPV infection, single-dose, KEN-SHE, Barnabas NEJM Evidence 2022); ACIP US considering.
• USPSTF 2024 breast final (Apr 30, 2024, JAMA 2024): mammography q2y ages 40 to 74 (Grade B), moved from age 50; addresses higher incidence in younger and Black women.
• GLP-1 agonists and cancer risk: GLP-1 receptor agonist (class) observational data (Wang JAMA Netw Open 2024) suggest reduced obesity-related cancers (uterine, colorectal, esophageal, HCC) vs insulin, with no benefit vs metformin; not tirzepatide-specific and hypothesis-generating; prospective trials ongoing.
• Wildfire smoke: IARC classified outdoor air pollution and PM as group 1 carcinogens (established cause of lung cancer; positive association only for bladder); residential wildfire exposure linked to modestly higher lung cancer and brain tumor incidence, no hematologic cancer association (Korsiak Lancet Planet Health 2022).
• Aspirin CRC prevention: USPSTF 2022 dropped CRC benefit from its aspirin recommendation (Grade C now for CVD prevention only, ages 40 to 59 with ≥10% ASCVD risk); not recommended for CRC prevention alone.
• H. pylori test-and-treat: community data support ~40% gastric cancer reduction in high-incidence populations; individualized in low-incidence US settings.

Disparities in cancer outcomes

• Black > White mortality for many cancers: multifactorial (access, biology, screening uptake). Black men have the highest overall incidence and mortality; Black women have low incidence but poor survival.
• Rural disparities: screening and specialist access.
• Socioeconomic / insurance status affects outcomes.

High-yield epi/prevention pearls

• Incidence is the best measure of prevention success; mortality depends on incidence plus survival; prevalence is not an occurrence measure.
• Lead-time and length-time bias and overdiagnosis can make a useless screen look beneficial; a demonstrated mortality reduction, ideally from a randomized trial, justifies population screening; stage shift or improved survival alone is insufficient.
• PPV rises with prevalence; sensitivity/specificity do not.
• Tobacco = #1 modifiable cause of cancer death; HPV vaccine prevents ~6 cancers.
• Lung: LDCT 50 to 80, 20 pack-years, current or quit ≤15 yr. Breast: start 40 (USPSTF 2024). CRC: start 45.
• Aspirin may reduce CRC (balance bleeding); H. pylori eradication reduces gastric cancer.
• Tamoxifen/AI reduce ER-positive breast cancer only; BRCA RRSO markedly reduces tubo-ovarian cancer (~80 to 96%); the breast-cancer benefit is uncertain and likely overestimated in older data, not a reliable ~50% reduction.
• Most vitamin/supplement chemoprevention trials are negative; some (beta-carotene, vitamin E, selenium) cause harm.