Breast Cancer (Part 1): Risk, Genetics, Screening, Diagnosis and Staging

Overview

• Epidemiology (US, 2025): ~310,000 new invasive cases + ~57,000 DCIS; ~42,000 deaths. Worldwide ~2.3M new cases in 2020, rising in economically developed regions (screening, reproductive factors, fat intake, BMI, age at menarche/lactation).
• Trends: incidence ↑ ~1%/yr, steeper (~1.4%/yr) in women <50 (↑ obesity, ↑ childbearing age); death rate ↓ ~1%/yr (better tx and detection).
• Most common non-skin ca in women; worldwide the most common cause of ca death in women, #2 cancer death in US women (after lung).
• Lifetime risk ~13% (1 in 8 women).
• 5-yr relative survival (AJCC stage): stage 0 ~99%, stage I ~99%, stage II ~93%, stage III ~72 to 75%, stage IV ~22 to 30%.
• Male breast cancer: <1% of breast ca; dx ~60 yrs; mostly HR+, often associated with BRCA2 or PALB2, Klinefelter, testosterone deficiency, or high estrogen-to-testosterone states (obesity, cirrhosis).

Risk factors

• Age and female sex are the most important risk factors.
• Incidence White > Black (↑ postmenopausal HRT use and ↑ screening). Mortality ~40% higher in Black women, who are more likely dx <40 yrs, at more advanced stage, with high-grade or triple-negative tumors.
• Incidence and mortality lower in Asian, Native American, and Hispanic populations.

• Alcohol: 1 drink/day associated with ~12% ↑ risk (dose-dependent).
• Obesity (BMI >30): ↑ risk in postmenopausal women.
• Smoking and low physical activity: modest ↑.
• Isoflavones (soy phytoestrogens), dairy, high-fat diets: unclear evidence.

• Early menarche (<12), late menopause (>55), nulliparity, late age at first pregnancy (>30) = extended estrogen exposure → ↑ risk for HR+ disease.
• Lactation and full-term pregnancy (terminal differentiation of breast epithelium) are protective.
• OCPs (both estrogen/progestin and progestin-only) ↑ risk, but absolute risk is very small.
• TNBC is associated with ↑ number of births and is NOT associated with nulliparity (opposite pattern to HR+).

• Combined estrogen + progestin → ↑ breast ca (HR ~1.26), especially HR+ and in >50 yrs; excess risk declines after d/c but persists for years (WHI 20-yr follow-up: CEE+MPA breast ca incidence HR 1.28). After the 2002 WHI report, US menopausal hormone use fell sharply; invasive breast ca incidence dropped 6.7% (2002 to 2003), 8.6% comparing 2001 with 2004, mainly in women ≥50 and ER+ tumors.
• Estrogen alone (post-hysterectomy): does NOT ↑ breast ca; may ↓.

• Thoracic RT at ages 10 to 30 (e.g., mantle RT for Hodgkin) → annual mammogram + annual breast MRI starting 8 yrs after RT but not before age 25 (whichever is later).

• Proportion of radiopaque (epithelial/stromal) vs radiolucent (fat) tissue. Categories: A (almost entirely fatty), B (scattered fibroglandular), C (heterogeneously dense, may obscure masses), D (extremely dense, lowers mammography sensitivity). C or D = dense.
• Exogenous hormones (HRT, OCP) ↑ density.
• DBT and contrast-enhanced mammography ↑ sensitivity in dense breasts. Supplemental modalities (whole-breast US, contrast-enhanced mammo, breast MRI, abbreviated MRI, molecular breast imaging) ↑ detection, but effect on outcomes is unknown; insurance coverage variable.
• Dense breast notification: federal mandate (effective Sep 10, 2024): mammography facilities must notify patients of breast density with standardized language (dense tissue ↓ mammographic sensitivity and ↑ risk; patient advised to discuss with their provider whether other imaging may help); supplemental imaging is not mandated.

• 5% to 10% associated with highly penetrant mutations (BRCA1, BRCA2). 15% to 20% associated with low-penetrance genes/SNPs and FHx.
• 1st-degree relative with breast ca → ~x2 risk. 1st-degree relative dx <50 or two 1st-degree relatives → ~x3 to x4 risk.

Genetic risk factors

• Proband: the affected individual through whom a family is first identified for genetic evaluation (index case); not necessarily the first relative tested.
• High-penetrance (RR >5) identifiable by pedigree; moderate-penetrance (RR 2 to 5) harder to identify (incomplete penetrance).
• Ashkenazi Jewish: ~2.5% carry 1 of 3 founder mutations (2 in BRCA1, 1 in BRCA2); these account for ~12% of breast ca and ~35% of ovarian ca in this population.
• BRCA1/2 are the main genes with treatment implications in breast ca (PARP inhibitors; olaparib also active in germline PALB2, TBCRC 048 ORR 82%).
• TP53 (Li-Fraumeni): ~90% lifetime malignancy risk, very-early-onset breast ca; ↑ risk of ca in RT field (avoid RT when possible); HER2+ more common.
• PTEN (Cowden): AD; breast with thyroid, endometrial, colon, renal; autism, macrocephaly, hamartomas, trichilemmomas.
• CDH1 (HDGC): diffuse gastric ca + invasive lobular breast ca; prophylactic total gastrectomy, usually ages 20 to 30 (individualized); endoscopic surveillance if deferred.
• PALB2 (partner and localizer of BRCA2): loss-of-function ↑↑ with FHx; breast risk near BRCA2 (mammo + MRI at 30, discuss RRM), but ovarian risk low (consider RRSO >45).
• RAD51C, RAD51D, BRIP1: RAD51C/D give ovarian ca and moderate breast ca risk (~20%, esp ER-/TNBC); BRIP1 gives ovarian ca with no established breast risk.
• BARD1 → TNBC. ATM and PALB2 → pancreatic ca. CHEK2 → ↑ breast ca (colorectal association modest, not established; average-risk CRC screening).
• ASCO-SSO 2024: offer BRCA1/2 testing to all newly diagnosed patients ≤65 yrs (and selected >65, e.g., PARP inhibitor candidates, TNBC); universal multigene panel testing regardless of family history is NOT yet standard of care.

Risk determination models

• Modified Gail model: most widely used; incorporates current age, age at menarche and first full-term pregnancy/nulliparity, number of biopsies and atypical hyperplasia, number of 1st-degree relatives, race. Good on a population basis but not robust for individuals; underestimates risk with strong genetic predisposition, and does NOT account for breast density or LCIS. Prevention strategies used if 5-yr risk >1.67%.
• Claus model: includes 1st- and 2nd-degree relatives with breast/ovarian ca and age at diagnosis.
• Tyrer-Cuzick (IBIS): comprehensive family history + BRCA + breast density. Preferred for high-risk assessment.
• BRCAPRO and BOADICEA: estimate probability of carrying a genetic mutation.
• NCCN criteria for genetic testing: refer if dx ≤50, TNBC at any age, male breast ca, ≥2 primaries, Ashkenazi ancestry, strong family history.

Prevention

• BRCA1 at 35 to 40, BRCA2 at 40 to 45 (onset ~8 to 10 yrs later), on completion of childbearing. Bilateral salpingectomy with delayed oophorectomy is an alternative but not SOC.
• Effects: ↓ ovarian ca ~85% (HR 0.14) and ovarian-specific mortality (HR 0.21); uncertain effect on first breast ca (older estimates BRCA1 HR 0.63, BRCA2 HR 0.36, Domchek 2010, likely biased; not confirmed prospectively, especially BRCA1); ↓ breast-specific mortality 55% (HR 0.44); ↓ all-cause mortality 60% (HR 0.40).
• Short-term menopausal hormone tx until ~age 50 if concern for early-menopause effects.

• ↓ breast ca risk >90% in hereditary breast/ovarian ca syndromes (BRCA1/2), mortality benefit not proven in RCTs, but observational data suggest reduced mortality, especially in BRCA1 carriers. Skin-sparing and nipple-sparing mastectomy are reasonable alternatives.

• SERMs ↓ invasive breast ca and ↓ fractures, but no effect on breast-specific or all-cause mortality.
• Tamoxifen x5 yrs (pre or postmenopausal): NSABP P-1, ↓ invasive breast ca ~50%. Adverse: ↑ endometrial ca (RR 3.28), ↑ VTE (DVT RR 1.44, PE RR 2.15; nonsignificant ↑ ischemic stroke (P-1); listed as a serious risk on label), cataracts, vasomotor symptoms, bone loss in premenopausal but ↓ osteoporotic fractures ~29% in women ≥50.
• Low-dose tamoxifen 5 mg/d x 3 yrs (TAM-01): 50% ↓ recurrence in ADH, LCIS, or DCIS with favorable toxicity.
• Raloxifene x5 yrs (postmenopausal only): ↓ ER+ invasive ca; inferior to tamoxifen (NSABP P-2 STAR) but ↓ endometrial ca and ↓ DVT/PE. Approved for prevention only (not treatment).
• Exemestane (postmenopausal): NCIC CTG MAP.3, ↓ invasive ca ~65% (~73% in ER+).
• Anastrozole (postmenopausal): IBIS-II, ↓ HR+ invasive ca and DCIS >50%.
• AIs have no endometrial ca or VTE risk but ↑↑ musculoskeletal symptoms and ↓↓ bone density.
• SERM meta-analysis (83,399 pts): 10-yr breast ca incidence 4.2% vs 6.3% (SERM vs placebo); no effect on ER-negative breast ca or overall mortality.
• After a breast ca dx, tamoxifen ↓ contralateral breast ca 42% (BRCA1) and 52% (BRCA2).

• Moderate exercise (2 to 3 h/wk) ↓ recurrence and all-cause mortality by ~40% to 67%. Low-fat diet: nonsignificant ~9% ↓ invasive breast ca (WHI DM, HR 0.91, CI 0.83 to 1.01); reduced breast ca mortality on long-term follow-up.

Screening

• USPSTF 2024: biennial mammography ages 40 to 74 (B recommendation; updated from starting at 50 in 2016).
• ACS (2015): annual 45 to 54, then biennial or annual ≥55 as long as life expectancy ≥10 yrs; option to start annual at 40.
• NCCN, ASCO, ACR: annual mammography starting at age 40.
• Meta-analyses: ~22% ↓ breast ca mortality (ages 39 to 74); others 31% to 49% (ages 40 to 75). Annual mammography beginning at age 40 prevents the most breast ca deaths. Harms: false-positive anxiety, unnecessary biopsies, overdiagnosis, ↑ DCIS detection.
• DBT: ↑ sensitivity and ↓ recall; dose comparable to 2D with synthetic reconstruction, roughly doubled if combined with separately acquired 2D. Breast self-exam and clinical breast exam are of no proven mortality benefit.

• MRI ↑ sensitivity to ~100% added to mammo, but combined imaging has lower specificity. MRI is not routine population screening but is risk-dependent: consider for dense breasts (ACR), and for LCIS/ADH or a personal history when lifetime risk is at least 20%.
• Indications: BRCA1/2 and other mutations (CHEK2, PALB2, CDH1, ATM, NF1, BARD1, Li-Fraumeni, Cowden); untested 1st-degree relative of a BRCA carrier; lifetime risk ≥20% (BRCAPRO or FHx-based models); chest RT at ages 10 to 30 (begin 8 yrs after RT, not before age 25).
• General rules: use gene-specific start ages (e.g., BRCA1/2 MRI at 25 with clinical exam q6 mo, mammo at 30; TP53 MRI at 20; ATM/CHEK2 MRI 30 to 35, mammo 40) (or earlier if breast ca dx <25). For an untested relative, start 10 yrs before the youngest family dx (but not later than 40).

• Screening (asymptomatic): CC and MLO views bilaterally ± tomosynthesis. Diagnostic (symptoms/abnormal exam/lesion): adds spot compression views.
• BI-RADS: 0 = incomplete (repeat imaging); 1 = negative (annual); 2 = benign (annual); 3 = probably benign, <2% malignancy (short-interval follow-up); 4 = suspicious, biopsy (4A ~5% to 4C ~95%); 5 = >95% malignancy, biopsy now; 6 = biopsy-proven malignancy.
• BI-RADS 4 and 5 (>2% malignancy) need biopsy; 6 = already biopsy-proven.

Diagnosis and prognostic indicators

• Palpable mass: US if <30 yrs, diagnostic mammography if >30 yrs. FNA acceptable for palpable axillary nodes or simple cysts; core needle superior to FNA for a breast mass (place radiopaque clip as locator).
• MRI: more sensitive, finds ~16% additional ipsilateral disease; ~8% convert BCT to mastectomy; ~11% need additional surgery after wide excision; inaccurate for noninvasive tumor size; contralateral MRI finds occult ca in ~3% (↑ biopsies and contralateral mastectomies without proven survival benefit).
• Nodes: macrometastasis >2 mm; micrometastasis >0.2 mm or >200 cells but none >2 mm; isolated tumor cells ≤0.2 mm or <200 cells (node-negative). Intramammary nodes count as axillary.
• LVI: ↓ prognosis, ↑ ipsilateral recurrence; does not preclude BCT but factors into RT.
• Tumor size: size of invasive component; DCIS size does not affect systemic risk (affects ipsilateral recurrence). Multifocal (≥2 foci in 1 quadrant) or multicentric (≥2 foci in separate quadrants) ↑ nodal positivity and ipsilateral recurrence; stage by largest focus (do not add sizes; use (m) modifier).

Pathology and biomarkers

• Invasive ductal (IDC, no special type): ~75% to 80%.
• Invasive lobular (ILC): ~10% to 15%. E-cadherin negative; vs IDC has ↑ age, ↑ size, ↑ multifocality, ↑ bilaterality, ↑ late recurrence, ↑ ER+, CDH1 inactivation, and ↓ HER2+, ↓ grade; harder to detect on mammo; mets to serosa, ovary, meninges, GI tract.
• Favorable variants: tubular and mucinous. Adenoid cystic: TNBC with MYB, indolent, no adjuvant chemo. Metaplastic: epithelial + mesenchymal, poor prognosis.
• Inflammatory breast ca (IBC): clinical dx (peau d'orange, erythema, edema); dermal lymphatic invasion. T4d. Trimodality mandatory.

• ER/PR (IHC): positive if ≥1% nuclear staining (ASCO/CAP). IHC measures ER-alpha (ESR1); PR is a marker of ER-alpha signaling. ER+ PR- is still HR+ and gets endocrine tx. HR+ associated with older age, lower grade, node-negative, ↑ DFS/OS; ~50% of HR+ recurrences occur in the first 5 yrs, with late recurrence up to 20 yrs. ~25% HR-; ~15% TNBC (highest recurrence in first 5 yrs).
• HER2 (IHC + ISH): EGFR family (EGFR-1, HER2, HER3, HER4); ~20% of breast ca; ↑ proliferation, higher grade, node+.
  • IHC 1+ or IHC 2+/ISH-negative = HER2-low; IHC 0 = HER2-zero (null or ultralow).
  • IHC 2+ = equivocal, reflex to ISH; if amplified, HER2-positive.
  • IHC 3+ or ISH-amplified = HER2-positive.
  • HER2-ultralow: IHC 0 with faint, incomplete membrane staining in ≤10% of cells. T-DXd eligible only if HR-positive metastatic disease that progressed on at least one endocrine therapy (DESTINY-Breast06).
• Ki-67: proliferation marker; ≥20% considered high (cutoffs vary; MIB-1 Dako Omnis used in monarchE). Higher Ki-67 = more chemosensitive.

• Luminal A (high ER/PR, low Ki-67, grade 1 to 2): best prognosis, endocrine-driven, chemo-insensitive.
• Luminal B (lower ER/PR, high Ki-67, grade 2 to 3): more aggressive, chemo-sensitive.
• HER2-enriched: HER2+, HR+ or HR-, high proliferation, grade 2 to 3.
• Basal-like / TNBC (ER/PR/HER2 negative): high proliferation, grade 3; basal markers CK5/6, EGFR.

Initial evaluation and workup

• Asymptomatic stage I to II: no systemic imaging needed.
• Stage III / locally advanced (T0 to 2 N2, T3 N1 to 2, any T4, or any N3): CT chest/abdomen ± pelvis + bone scan. Stage II to III: PET can evaluate regional nodes and distant sites at once, with fewer false positives than CT + bone scan (allows earlier neoadjuvant start).
• Genetic counseling for FHx; germline testing per ASCO-SSO 2024 (BRCA1/2 for all newly diagnosed ≤65 yrs; selected >65) and NCCN criteria. Nonhormonal contraception; refer to gynecology.
• Fertility preservation: oocyte, sperm, embryo cryopreservation. GnRH agonist (goserelin) starting ~1 week before chemo (POEMS, in HR- disease) ↑ successful pregnancy and ↓ premature ovarian failure, but is not a substitute for established fertility preservation. Chemo can be delayed up to ~6 weeks. Embryo fertilization needs a partner/donor.

Staging (AJCC 8th ed, anatomic + prognostic)

• Tis: DCIS or Paget of nipple without invasive ca/DCIS. LCIS removed from staging (benign entity).
• T1 ≤20 mm (T1mi ≤1 mm; T1a >1 to 5 mm; T1b >5 to 10 mm; T1c >10 to 20 mm). T2 >20 to 50 mm. T3 >50 mm. T4: T4a chest wall (not pectoralis alone), T4b skin ulceration/satellite nodules/edema (peau d'orange), T4c both, T4d inflammatory.
• Invasion of dermis alone does not qualify as T4. IBC requires typical skin changes over ≥1/3 of the breast (dermal lymphatic invasion is supportive, not required or sufficient alone).
• M: M0 vs M1.
• Prognostic stage groups incorporate grade, ER, PR, HER2, and multigene assays; e.g., a T2N0M0 (anatomic IIA) ER+/HER2- tumor with Oncotype RS <11 is downstaged to prognostic Stage IA. Presence of ca in vessels, lymphatics, or nodes precludes classification as a pathologic complete response.

Gene expression signatures

• Classifies tumors as high or low genomic risk. MINDACT enrolled up to 3 node+ (~80% N0, ~20% N1); assessed both clinical risk (Adjuvant! Online: T stage, N, grade) and genomic risk (MammaPrint).
• Identifies a subset with low likelihood of distant recurrence despite high-risk clinical features. Low/low: no chemo. High/high: chemo. High clinical/low genomic: slight benefit from chemo. Low clinical/high genomic: no benefit from chemo.
• Use MammaPrint only in clinically high-risk patients.

• Defines recurrence risk in ER+ breast ca treated with tamoxifen x5 yrs. Classic categories: low (RS <18), intermediate (RS 18 to 30), high (RS ≥31). TAILORx and RxPONDER refined treatment cutoffs (below).
• Chemo for sure: RS ≥26 in pre- or postmenopausal regardless of N; or N1 in premenopausal regardless of RS.
• No chemo: RS <26 + postmenopausal (N0 or 1 to 3 nodes), or RS <16 in premenopausal with pN0.
• Consider chemo vs OFS + tamoxifen/AI: premenopausal, RS 16 to 25, pN0.
• Give chemo for any N2 (≥4 nodes); do NOT use RS in N2. Oncotype is not used in DCIS.
• TAILORx (N0): randomized RS 11 to 25; RS 0 to 10 endocrine alone, RS 26 to 100 assigned chemoendocrine. Endocrine alone for age >50 with RS ≤25; chemo benefit in age ≤50 with RS 16 to 25. RxPONDER (1 to 3 node+): postmenopausal RS 0 to 25 do NOT benefit from chemo; premenopausal RS 0 to 25 DO benefit (likely a chemo-induced ovarian suppression effect).

• Breast Cancer Index (BCI): predicts late recurrence and benefit from extended endocrine tx. Low BCI (0 to 5) in T1-2 N0 HR+ HER2-: low late recurrence, less benefit from extension beyond 5 yrs. High prognostic BCI: higher late recurrence. Benefit from extended endocrine therapy is predicted by BCI(H/I)-High, not by the prognostic score alone.
• Prosigna (PAM50): intrinsic subtype + ROR score; FDA-approved for postmenopausal HR+ N0. EndoPredict also validated for late recurrence.

Continued in: Breast Cancer (Part 2): DCIS, Early-Stage and Metastatic Treatment