[Syllabus]: Breast Cancer Screening in the era of Tomosynthesis

Introduction

This presentation provides a comprehensive review of breast cancer screening in the era of digital breast tomosynthesis (DBT). It covers breast anatomy and oncologic pathophysiology, epidemiology and risk factors, the evidence base for screening mammography including trial data and inherent biases, performance metrics and screening intervals, benefits and harms (including radiation considerations), the role and limitations of clinical and self-examination, MQSA-driven workflow, the technology and clinical performance of DBT, breast density and legislation, supplemental screening considerations in dense breasts, and international screening strategies alongside current U.S. controversies.

Breast Anatomy and Oncologic Pathophysiology

Breast tissue comprises skin, subcutaneous fat, and glandular elements organized into lobules and ducts that converge at the nipple. Most breast cancers arise in the terminal ductal lobular unit (TDLU). Histologic progression spans benign epithelial proliferation to atypia, ductal carcinoma in situ (DCIS), microinvasion, and invasive carcinoma.

Key Points

• ~80% of breast cancers are ductal; ~10% lobular; remainder arise from stroma, fat, or skin (rare).
• DCIS is confined within ducts; invasion is defined by basement membrane breach (microinvasive ≤1–2 mm).
• Invasive lobular carcinoma may be less radiodense and harder to detect on mammography.

Epidemiology and Survival

Breast cancer is the most common cancer in U.S. women; lung cancer remains the leading cause of cancer mortality. Incidence has modestly increased recently, yet mortality has fallen substantially since the late 1980s, reflecting the combined impact of screening and improved therapies.

Key Points

• 2019 estimates: ~268,000 new invasive cases; ~62,000 DCIS; ~41,000 deaths.
• Mortality decreased ~40% from 1989 to 2016; >3.1 million survivors in the U.S.
• Peak diagnosis: ages 60–64; ~5% of cases occur <40 years.
• Lifetime risk ≈12.4% (~1 in 8); “high risk” commonly ≥20%.
• Racial differences: incidence in older non-Hispanic Black women is lower than in White women, yet mortality is higher.
• Five-year survival: overall ~90%; localized disease >99%; distant metastasis ~22%.

Risk Factors for Breast Cancer

Risk is driven largely by non-modifiable factors. While many exposures are discussed by patients, several do not increase risk based on current evidence.

Key Points

• Strong, non-modifiable factors: age; family history (e.g., two or more first-degree relatives); prior early-onset breast cancer; breast density.
• Modifiable factors (e.g., alcohol use, anthropometrics, reproductive history) have comparatively small effects.
• About 10% of breast cancers are linked to identifiable genetic mutations.
• No demonstrated risk increase: breast implants, abortion, bra type, hair dye.
• Breast density is both a risk factor and a determinant of mammographic sensitivity.

Goals and Principles of Screening

Screening aims to identify asymptomatic disease earlier to enable curative intervention, shift stage at diagnosis, and reduce mortality. Programmatic screening must account for tumor biology and test performance.

Key Points

• Screening seeks earlier detection (stage shift) and mortality reduction, not just improved survival statistics susceptible to bias.
• Interval cancers reflect biological growth between routine screens.
• Screening policy should consider population risk, tumor sojourn time, and program feasibility.

Evidence Base: Randomized Trials, Biases, and Mortality Impact

Multiple randomized controlled trials (RCTs) and meta-analyses have evaluated screening mammography. Understanding biases is critical to interpreting outcomes.

Key Points

• Large RCTs (U.S., Europe, Canada): meta-analysis of seven trials shows ~15% mortality reduction; one Canadian trial reported no benefit.
• Stage shift: greater detection of node-negative, stage 0–I cancers in screened groups.
• Key biases:

- Lead-time bias: earlier diagnosis inflates survival time without prolonging life.

- Length/sojourn-time bias: screening preferentially detects slower-growing tumors.

- Selection bias: healthier individuals more likely to participate in screening.

- Overdiagnosis: increased detection of DCIS and indolent lesions that may not cause death.

• Long follow-up (often 15–20 years) is required to assess mortality effects.

Screening Interval and Age Considerations

Optimal screening frequency reflects estimates of tumor sojourn time and balances benefits against harms (e.g., false positives).

Key Points

• Estimated mean sojourn time ~3 years (U.S. ~2.5, Canada ~3, Europe ~4.3 years).
• Annual screening may not capture very fast-growing tumors; biennial may suffice for slower-growing disease.
• Evidence in women 40–49 is less robust due to smaller enrolled cohorts and statistical power limitations in early trials.
• Data are limited for women >75 due to under-enrollment; decision-making should consider life expectancy (~13.5 years at age 75) and willingness for treatment.

Performance Metrics and Outcomes in Practice

Program-level metrics inform quality and expectations for screening mammography.

Key Points

• From large U.S. cohorts (e.g., BCSC):

- Detection rate: ~4.4 cancers per 1,000 screening mammograms.

- Recall rate: ~9.7% (~10%).

- DCIS proportion among screen-detected cancers: ~21%.

- 76% of screen-detected cancers are stage 0–I; mean size ~15.9 mm.

- Cumulative false-positive callback after 10 annual screens: ~49%.

• Sensitivity ~80% (misses ~20% of cancers).
• Interval cancer rate: ~1–3 per 1,000 women between routine screens.
• Typical 1,000-screen workflow: ~100 recalls; ~20 biopsies; ~5 cancers diagnosed.

Benefits and Harms of Screening Mammography

Screening confers mortality reduction but entails trade-offs including overdiagnosis, false positives, and radiation exposure.

Key Points

• Benefits: earlier stage at detection, mortality reduction demonstrated in most RCTs.
• Harms:

- Overdiagnosis (especially increased DCIS detection).

- False positives leading to additional imaging/biopsies and anxiety.

- False reassurance if normal screen delays evaluation of symptoms.

- Radiation exposure: approximately 1.8 mGy per view; four views ~7 mGy.

• Dose context:

- Annual background radiation ≈3 mSv; a mammogram approximates ~7–8 weeks of background exposure.

- Simulation (40–74 years, 100,000 women): up to 16 radiation-induced deaths vs ~968 deaths averted via early detection—benefit outweighs risk.

• Hormone therapy reduction around 2000 associated with decreased incidence thereafter.

Clinical Breast Exam and Self-Exam

The independent value of clinical breast examination (CBE) and breast self-examination (BSE) remains uncertain in the context of population screening.

Key Points

• No large RCT has isolated the independent benefit of CBE within organized screening programs.
• Ongoing studies (e.g., India, Egypt) are evaluating CBE where mammography access is limited.
• BSE has not demonstrated mortality benefit.

Regulatory and Workflow Considerations

Quality in screening mammography is governed by federal regulation and best practices in image acquisition and interpretation.

Key Points

• Dedicated mammographic equipment and skilled technologists are required; MQSA (1992) regulates quality.
• Batch reading (40–50 exams over ~2 hours) and maintaining recall rate ~10% are common practice targets.
• Program metrics guide continuous quality improvement.

Digital Breast Tomosynthesis: Technology and Dose

DBT is a quasi-3D mammographic technique that acquires multiple low-dose projections over a limited arc and reconstructs slice images.

Key Points

• FDA approved first DBT system in 2011.
• Radiation dose per DBT view approximates a single 2D view; total dose depends on whether synthesized 2D images replace separate 2D acquisitions.

- With synthesized 2D: dose similar to standard 2D.

- With separate 2D plus DBT: dose per view is roughly doubled.

• Compression remains necessary:

- Reduces tissue thickness and dose.

- Minimizes motion during longer DBT acquisition to preserve image quality.

Digital Breast Tomosynthesis: Clinical Performance, Indications, and Limitations

DBT improves detection and reduces recalls in many settings but does not eliminate mammography’s fundamental constraints.

Key Points

• In population studies, DBT (with 2D) detects ~1–2 additional cancers per 1,000 screens and lowers callback rates.
• Greatest benefit:

- Heterogeneously dense breasts.

- Masses and architectural distortion.

• Limited or marginal benefit:

- Calcifications (evidence mixed).

- Almost entirely fatty breasts (little added value).

- Extremely dense breasts (both 2D and DBT may miss cancers).

• Persistent limitations:

- Density-based contrast; limited sensitivity for lesions without increased radiodensity (e.g., invasive lobular carcinoma).

- Field-of-view constraints; tissue outside standard views remains unassessed.

• Unknown whether DBT justifies lengthening screening intervals; interval strategy is still driven by tumor biology.

Breast Density: Risk, Detection Challenges, and Legislation

Breast density affects both cancer risk and mammographic sensitivity. Reporting is mandated nationally.

Key Points

• BI-RADS density categories:

- a: Almost entirely fatty.

- b: Scattered fibroglandular densities.

- c: Heterogeneously dense.

- d: Extremely dense.

• “Dense” = categories c and d; ~43% of U.S. women 40–74 have dense breasts.
• Extremely dense breasts confer ~4–6× higher risk vs fatty breasts; most patients fall between extremes.
• Higher density reduces mammographic sensitivity; DBT partially mitigates but does not eliminate this issue.
• Legislation:

- State density notification laws in 38 states led to national MQSA amendment (March 2019) mandating patient notification of density.

• Communication gap persists: patients are informed of density-related limitations, but evidence-based guidance for adjunct testing is uncertain.

Supplemental Screening in Dense Breasts

Whether to add supplemental modalities for women with dense breasts remains unresolved at a population level.

Key Points

• USPSTF: current evidence insufficient to recommend for or against supplemental screening in women with dense breasts.
• Programmatic implications: discuss individualized risk; acknowledge uncertainty; DBT may provide incremental benefit in many dense breasts.
• Approximately 43% of screening-eligible women are affected by density-related considerations.

International Screening Programs and U.S. Controversies

Screening intervals vary internationally, and U.S. recommendations reflect differing interpretations of evidence, especially in younger and older age groups.

Key Points

• International examples:

- Canada/Australia/New Zealand: biennial, ages 50–69.

- Sweden: 40–49 every 18 months; 50–69 biennial.

- United Kingdom: triennial starting at 50.

• U.S. controversies:

- Benefit in ages 40–49 less statistically robust in early U.S. RCTs due to limited power; debate persists regarding start age and interval.

- Limited trial data >75 years due to under-enrollment; decisions guided by life expectancy and treatment candidacy.

- A Canadian RCT reporting no benefit has been criticized for methodological limitations.

Conclusion

Breast cancer screening has reduced mortality through earlier detection and stage shift, supported by multiple randomized trials despite recognized biases. Contemporary practice demands nuanced balancing of benefits and harms, informed by performance metrics, patient age, tumor biology, and breast density. Digital breast tomosynthesis improves detection and decreases recalls—especially in heterogeneously dense breasts—while retaining mammography’s core limitations and not yet redefining optimal intervals. Density notification is now federally mandated, yet robust guidance for supplemental screening remains insufficient. Ongoing research, careful communication, and individualized, evidence-informed screening strategies are essential to optimize outcomes in the era of tomosynthesis.