Swedish Breast Database Study

Survival After Breast Conservation vs Mastectomy Adjusted for Comorbidity and Socioeconomic Status

de Boniface et al., JAMA Surgery 2021

Background

• Breast-conserving surgery (BCS) with radiotherapy (RT) is equivalent to mastectomy for early breast cancer
• Recent studies suggest improved survival with BCS+RT over mastectomy
• SweDCIS trial: RT after BCS for DCIS reduced ipsilateral events
• Swedish Breast Cancer Group 91 trial: Benefit of RT after BCS in all subtypes, especially with low TILs
• Comorbidity and socioeconomic status may confound survival outcomes

Study Objective

To determine whether the reported survival benefit of breast conservation is eliminated by adjustment for comorbidity and socioeconomic status

Study Design

• Cohort study using Swedish national registers
• Data from 2008 to 2017
• Primary invasive T1-2 N0-2 breast cancer patients
• Three treatment groups:
  • BCS with RT (BCS+RT)
  • Mastectomy without RT (Mx-RT)
  • Mastectomy with RT (Mx+RT)

Patient Population

Inclusion Criteria:

• Women diagnosed with primary invasive T1-2 N0-2 breast cancer
• Underwent breast surgery in Sweden from 2008 to 2017
• Clear microscopic resection margins

Exclusion Criteria:

• Previous malignancy (except basal cell skin cancer and cervical intraepithelial neoplasia)
• Ipsilateral breast implant
• Concurrent chemoradiotherapy

Patient Characteristics

[Insert table of patient characteristics]

• Total patients: 48,986
• BCS+RT: 29,367 (59.9%)
• Mx-RT: 12,413 (25.3%)
• Mx+RT: 7,206 (14.7%)

Methods

Data Sources:

• National Breast Cancer Quality Register (NKBC)
• Patient Registers for comorbidity data
• Statistics Sweden for socioeconomic data

Comorbidity Assessment:

• Charlson Comorbidity Index (CCI)
• Assessed within 12 months before treatment

Statistical Methods

• Primary outcomes: Overall survival (OS) and breast cancer-specific survival (BCSS)
• Kaplan-Meier estimates and Cox proportional hazards models
• Stepwise adjustment for confounders:
  • Model 1: Age, year, region
  • Model 2: + Tumor characteristics
  • Model 3: + Socioeconomic factors
  • Model 4: + Charlson Comorbidity Index
• Subgroup analyses by prognostic groups and follow-up time

Results: Overall Survival

[Insert Kaplan-Meier curve for overall survival]

• 5-year OS: 91.1% (95% CI, 90.8-91.3)
• BCS+RT showed better OS compared to Mx-RT and Mx+RT
• Adjusted HR for Mx-RT vs BCS+RT: 1.79 (95% CI, 1.66-1.92)
• Adjusted HR for Mx+RT vs BCS+RT: 1.24 (95% CI, 1.13-1.37)

Results: Breast Cancer-Specific Survival

[Insert Kaplan-Meier curve for BCSS]

• 5-year BCSS: 96.3% (95% CI, 96.1-96.4)
• BCS+RT showed better BCSS compared to Mx-RT and Mx+RT
• Adjusted HR for Mx-RT vs BCS+RT: 1.66 (95% CI, 1.45-1.90)
• Adjusted HR for Mx+RT vs BCS+RT: 1.26 (95% CI, 1.08-1.46)

Subgroup Analysis: Prognostic Groups

[Insert forest plot of HRs by prognostic groups]

• T1N0: Mx-RT and Mx+RT associated with worse OS and BCSS
• T1N1: Mx-RT associated with worse OS; no difference in BCSS
• T2N0: Mx-RT and Mx+RT associated with worse OS and BCSS
• T2N1: Mx-RT associated with worse OS and BCSS; Mx+RT only worse for OS
• T1N2 and T2N2: Mx-RT associated with worse outcomes; no significant difference for Mx+RT

Subgroup Analysis: Follow-up Time

• Short-term follow-up (0-5 years):
  • Similar associations for OS across all groups
  • Stronger associations for BCSS in Mx groups
• Long-term follow-up (>5 years):
  • Persistent associations for OS
  • Some prognostic groups showed stronger BCSS associations for Mx+RT

Detailed Benefits of BCS+RT

• Improved overall survival across all prognostic groups
• Better breast cancer-specific survival in most subgroups
• Benefit persists after adjusting for comorbidity and socioeconomic factors
• Consistent advantage in both short-term and long-term follow-up
• Particularly strong benefit in node-negative patients (T1N0, T2N0)
• Reduced risk of overtreatment in low-risk patients

Impact of Comorbidity and Socioeconomic Status

• Adjustment for CCI did not substantially alter survival differences
• Socioeconomic factors (education, income) had minimal impact on results
• Mx-RT group had lower education levels and lower family income
• Both Mx groups had higher comorbidity burden than BCS+RT
• Survival benefit of BCS+RT remained significant despite adjustments

Critique of Study Results

• Contradicts randomized trials showing no OS difference between mastectomy and BCS
• Potential limitations in stage adjustment:
  • Adjusted by T and N stage, not exact tumor size or node number
  • Mastectomy group may have had larger T2 tumors (e.g., 4.9 cm vs 2.1 cm)
  • N1/N2 stages may have had more involved nodes in mastectomy group
• Possible mechanisms for BCS+RT benefit:
  • RT-induced immune response
  • RT targeting occult disease missed by mastectomy
• Large sample size may detect differences not seen in smaller trials
• Unmeasured confounders:
  • Smoking, body habitus, alcohol use, drug abuse, psychiatric disorders
  • These factors might actually favor the mastectomy group
• Selection bias: BCS patients may be healthier overall
• Evolving patterns of care over the study period (2008-2017)

Discussion

• BCS+RT showed better survival outcomes than mastectomy, with or without RT
• Survival benefit persisted after adjusting for comorbidity and socioeconomic status
• Results contradict randomized trials showing no OS difference
• Potential limitations in stage adjustment:
  • Mastectomy group may have had larger tumors within T stages
  • More extensive nodal involvement within N stages
• Possible explanations for BCS+RT benefit:
  • Immune response generated by RT
  • Targeting of occult disease missed by mastectomy
  • More precise staging in BCS patients
• Large sample size may detect differences not seen in smaller trials
• Unmeasured confounders may impact results
• Selection bias: BCS patients may be healthier overall

Summary

• BCS+RT associated with better OS and BCSS compared to mastectomy
• Survival benefit not explained by comorbidity or socioeconomic factors
• Benefit consistent across prognostic groups and follow-up periods
• Results support the use of BCS+RT when both options are valid
• Mastectomy should not be considered equivalent to breast conservation
• Further research needed to understand mechanisms behind survival differences