396 – Breast cancer screening: understanding risk, deciding when to start, and more

KEYPOINTS

• 01

  Screening works, but under-screening is the fixable reason women still die

  About 1 in 8 US women develop invasive breast cancer and roughly 42,000 die each year. Stage 1 disease has over 96% ten-year survival versus about 30% five-year survival at stage 4, and regular screeners are up to 40% less likely to die of it. Peter estimates only ~7–10% of cases are biologically aggressive enough to evade even perfect screening; the bigger, solvable problem is that roughly a third of women over 40 haven't had a mammogram in two years and ~20% of 50–74-year-olds aren't up to date.

• 02

  The MRI gap is a pure execution failure

  At least 9% of women meet guideline criteria for screening MRI, yet actual utilization is about 0.4%. Peter stresses MRI isn't controversial for these women — the high-risk group is identifiable and the tool exists; the system simply fails to connect the two, partly because insurance follows USPSTF guidance that doesn't address high-risk women.

• 03

  Guidelines disagree because they answer a population question

  The composite from the American Cancer Society, NCCN and American College of Radiology: formal risk assessment by about 25, annual mammography from 40 for average risk, earlier MRI for high risk, and screening as long as you'd accept treatment. The outlier is the USPSTF — biennial mammography for ages 40–74 — which balances cost and false positives across millions. Peter's framework for individuals: know baseline risk, decide your false-positive tolerance, then pick modality and frequency.

• 04

  Risk is usually many small inputs, not one red flag

  Beyond sex (1 in 8 women versus about 1 in 750 men) and age (median diagnosis ~62), factors include BRCA1/2 (only ~1 in 400 people, more in Ashkenazi Jewish ancestry), family history (which also captures low-penetrance variants and can show up as prostate or pancreatic cancer), ancestry (Black women diagnosed younger with more aggressive subtypes), prior high-dose chest radiation such as adolescent Hodgkin's treatment, density, reproductive factors and modifiable ones like alcohol, obesity and inactivity. Calculators such as Tyrer-Cuzick combine these; lifetime risk above 20% is generally 'high risk'.

• 05

  Breast density is both a risk factor and a blind spot

  Dense tissue modestly raises risk and makes mammograms harder to read because tumors and dense tissue both appear white. About half of screening-age women are dense (BI-RADS C or D), and the FDA now requires centers to notify women. The catch-22 is you usually learn your density only once screening starts, but it is roughly 60–70% heritable, so a mother's or grandmother's density is worth asking about.

• 06

  False positives are real but mostly cost anxiety

  About 10% of US screening mammograms trigger a callback, and only ~5% of callbacks end in cancer — roughly 95 of every 100 recalled women don't have it. More than half of women screened annually for 10 years get at least one false positive. The higher your baseline risk, the easier it is to justify that burden; for average-risk women it's a preference call.

• 07

  There is a clear hierarchy of imaging tools

  Mammography — ideally 3D tomosynthesis (DBT), which detects more cancer with fewer recalls, especially in dense breasts — is the foundation and is best at calcifications like DCIS (25–60% of which may become invasive). MRI with gadolinium is the most sensitive tool and supplements rather than replaces mammography. Contrast-enhanced mammography is the backup when MRI isn't feasible. Ultrasound adds the least consistent value and is the most operator-dependent.

• 08

  Abbreviated MRI is the most underused tool

  Peter says the abbreviated protocol keeps nearly all of full MRI's sensitivity in 10–15 minutes instead of 30–60, making it cheaper and more scalable. In women with extremely dense breasts, adding MRI after a negative mammogram halved interval cancers from 5 to 2.5 per 1,000.

• 09

  Screen annually if you're optimizing for your own survival

  No RCT compares annual versus biennial screening on mortality; it rests on CISNET modeling. The 2009 analysis (film mammography) found biennial kept ~81% of the benefit with half the false positives — an efficiency argument. The 2024 CISNET analysis with digital and DBT showed annual screening from 40–79 cut mortality 42% versus 30% biennial. Per exam, false positives were actually lowest with annual screening, likely because radiologists have a recent comparison image.

• 10

  Starting age depends on risk more than the calendar

  Only ~5% of diagnoses occur under 40 and cumulative risk to 40 is under 1%; registry incidence rises smoothly with no spike at 40, so low early incidence looks real. But younger women get more aggressive disease — about 20% triple-negative under 40 versus 6–12% after — and triple-negative tumors can double in under four months.

• 11

  Inflammatory breast cancer and symptoms bypass the screening logic

  Inflammatory breast cancer (1–5% of cases) often has no lump — instead swelling, redness, warmth or skin thickening — and may not show on a mammogram, so a normal screen doesn't rule it out. Any new symptom, in women or men, warrants prompt diagnostic evaluation rather than waiting for the next screen.