Introduction
This presentation reviews contemporary surveillance strategies for individuals with a personal history of breast cancer, emphasizing evidence-based imaging options, evolving performance data, and areas of ongoing controversy. Topics include the rationale for intensified follow-up, comparative guidance across professional societies, the roles of mammography, digital breast tomosynthesis (DBT), breast MRI, contrast-enhanced mammography (CEM), and whole-breast ultrasound, as well as nuanced issues such as imaging reconstructed breasts, when to discontinue screening, screening versus diagnostic pathways, live versus batch interpretation, and improving adherence through risk-based, patient-centered approaches.
Clinical Rationale for Surveillance After Breast Cancer
Individuals with a personal history of breast cancer face an elevated risk of recurrence or new primary breast cancers compared with the general population. Post-treatment mammographic changes can complicate interpretation, especially within the first two years after therapy. As survival improves due to screening, earlier stage detection, and effective systemic therapy, optimizing long-term surveillance becomes increasingly important.
Key Points
- Relative risk for recurrence or new cancer: approximately 2.4–4.0, peaking in the second year post-treatment
- Post-treatment changes (e.g., architectural distortion, edema, skin thickening) often plateau by two years but may persist
- Benefits: earlier detection, less invasive therapy, improved survival; Risks: false-positive recalls/biopsies
- Interval cancer rates are about fourfold higher than in average-risk screening populations
Current Surveillance Guidelines and Variability
Professional society recommendations vary in initiation and supplemental imaging but generally endorse annual mammography. MRI usage criteria differ, particularly regarding age at diagnosis, breast density, and calculated risk for second primaries.
Key Points
- Annual mammography recommended by ASCO, NCCN, ACS, SBI/ACR
- Initiation timing varies (e.g., 1 year after first post-treatment mammogram and ≥6 months post-radiation vs no specified timing)
- ASCO: MRI for lifetime risk >20% (e.g., BRCA1/2, strong family history)
- SBI/ACR: consider annual MRI (with or without contrast) for personal history with dense breasts or diagnosis before age 50
- Persistent heterogeneity in recommendations for MRI frequency and initiation
Breast MRI in Surveillance
MRI provides high sensitivity for detecting clinically occult disease in the post-treatment breast and contralateral breast. Performance appears sustained across multiple screening rounds with both standard and abbreviated protocols, with notably low interval cancer rates.
Key Points
- Reported cancer detection rate: 1.9–58.9 per 1,000 examinations; sensitivity: 61.4–100%; biopsy PPV: ~10–43.5%
- Low interval cancer rates with MRI (~1.5/1,000) vs mammography (up to ~3.5/1,000)
- Detected cancers often small, node-negative; more frequent in moderate–marked BPE and hormone receptor–negative disease
- Emerging data suggest second-cancer risk in personal-history patients may approximate BRCA carriers, arguing for intensified intervals in selected patients
- Abbreviated MRI protocols can sustain detection performance across rounds
Background Parenchymal Enhancement as a Risk Marker
Background parenchymal enhancement (BPE) on contrast-enhanced MRI reflects normal tissue vascularity and has emerged as an independent risk marker in surveillance populations.
Key Points
- Any BPE (mild, moderate, marked) is associated with increased subsequent cancer risk on surveillance MRI
- BPE should inform risk stratification and may influence the decision to include supplemental MRI in surveillance
Digital Breast Tomosynthesis (DBT)
DBT reduces false-positive recalls while maintaining cancer detection rates and biopsy PPV relative to full-field digital mammography, mirroring trends in average-risk screening and offering particular value in complex post-treatment breasts.
Key Points
- Decreases abnormal interpretation (recall) rates versus 2D mammography
- Maintains cancer detection rates and PPV for biopsy
- Utility demonstrated for architectural distortion and subtle asymmetries detectable only on tomosynthesis
- Further analyses needed across age, density, and prior imaging availability
Contrast-Enhanced Mammography (CEM)
CEM leverages iodinated contrast to depict tumor vascularity, offering a practical alternative when MRI is contraindicated or unavailable. Early surveillance data are promising.
Key Points
- Diagnostic sensitivity comparable to MRI in several settings; in surveillance, early studies show higher sensitivity than mammography alone
- Study of 1,191 women with prior breast cancer: recall 6.1%, PPV 47.9%, 55% reduction in screening ultrasound utilization
- May be particularly helpful in patients with MRI contraindications or limited access
- Impact across multiple screening rounds remains under investigation
Whole-Breast Ultrasound
While ultrasound can detect mammographically occult lesions, especially in dense breasts, it carries higher recall rates and lower PPV compared with mammography.
Key Points
- ACRIN 6666: additional cancer yield similar in those with and without prior cancer (~4.2–4.3/1,000); sensitivity ~28.8%
- Increased recalls and lower PPV are key trade-offs
- Considered a potential alternative for high-risk patients who cannot undergo MRI (per SBI/ACR)
- Best positioned as a selective adjunct rather than a universal surveillance tool
Surveillance of the Reconstructed or Post-Mastectomy Breast
Guidance diverges regarding routine imaging after mastectomy with or without reconstruction. Most recurrences are clinically detected, and pooled detection rates for imaging are lower than typical screening benchmarks.
Key Points
- NCCN/ASCO/ACS: recommend against routine imaging post-mastectomy with or without reconstruction
- ACR: mammography may be appropriate for autologous reconstructions (with or without implants); not indicated without reconstruction
- Local recurrence rates: ~2–7.5% after modified radical mastectomy; ~2–4% in reconstructed breasts
- Meta-analysis (16 studies): pooled detection rates lower than BI-RADS benchmarks (mammography 1.86 vs 4.7; ultrasound 2.66 vs 3.7; MRI 5.17 vs 20 per 1,000), supporting no routine imaging surveillance
- Clinical exam remains paramount; when imaging is performed, consider anastomosis “contact zone” as a common recurrence site and differentiate from fat necrosis
Determining When to Discontinue Surveillance
No universal age cutoff exists. Decisions should weigh residual risk, life expectancy, tumor biology, prior treatment, comorbidity, and patient preferences.
Key Points
- Increasing age raises baseline risk, but competing mortality and treatment preferences matter
- International Society of Geriatric Oncology: consider discontinuation when life expectancy <5 years
- Use decision aids and shared decision-making to align with patient goals
Screening vs Diagnostic Pathways After Treatment
Whether to categorize asymptomatic surveillance as screening or diagnostic varies by institution, influencing access, out-of-pocket costs, and operational logistics.
Key Points
- Many practices perform diagnostic mammography at 6–12 months post-treatment, often extending through 2 years; a subset continues 2–5 years
- Insurance and scheduling differences favor screening pathways for access and cost
- Data from 707 women: most additional views were needed in year 1, supporting return to screening by year 2
- Recent institutional evidence suggests immediate return to screening may be feasible post-treatment
Live (Real-Time) vs Batch (Offline) Interpretation
Interpretation workflow impacts patient experience and clinic operations. Performance metrics are generally preserved across approaches, with nuanced differences in recall and equity considerations.
Key Points
- Live interpretation: reduces time to diagnostic work-up and patient anxiety; may improve access for some underserved groups; operationally unpredictable
- Batch/offline: requires separate diagnostic appointments; may reduce abnormal interpretation (callback) rates
- Cancer detection rates and PPV are preserved across approaches; downstream morbidity/mortality impact remains unproven
- Balance performance, access, and resource constraints when selecting workflow
Improving Adherence and Equity in Surveillance
Adherence diminishes over time, with disparities by age, race/ethnicity, education, and income. Targeted strategies can mitigate gaps.
Key Points
- First-year adherence: ~63–86%; declines thereafter
- Lower adherence among younger patients; Black and Hispanic patients; individuals with lower education or income
- Interventions: structured survivorship plans, patient education, expanded clinic access, navigation and social support, insurance facilitation
- Proactive outreach to identified at-risk groups is essential
Toward Risk-Based, Personalized Surveillance and the Role of AI
Personalized surveillance integrates clinical, imaging, and social determinants to tailor modality and interval, with AI models offering potential augmentation pending robust validation.
Key Points
- Inputs: tumor subtype, age, prior therapies, breast density, family/genetic risk, and socioeconomic factors (income, education, insurance)
- Goals: align modality selection (e.g., MRI, DBT, CEM, ultrasound) and interval with individualized risk and preferences
- AI/machine learning may refine second-cancer risk stratification; requires diverse, representative validation and attention to implementation barriers
- Emphasize shared decision-making and equity to ensure clinical adoption benefits all subgroups
Conclusion
Surveillance for patients with a personal history of breast cancer should be anchored by annual mammography, with selective use of DBT, MRI, CEM, and ultrasound based on individualized risk, access, and patient preferences. MRI and CEM enhance detection of clinically occult disease; DBT reduces recalls while maintaining key performance metrics. Routine imaging of the post-mastectomy breast remains unsupported by current evidence, and discontinuation of surveillance should be guided by life expectancy and shared decision-making. Operational choices (screening vs diagnostic pathways; live vs batch interpretation) should balance accuracy, access, cost, and equity. Future progress hinges on robust risk-based models—including AI—to personalize surveillance and improve adherence across diverse populations.
