Exploring Prognostic Factors in Breast Cancer Survival: An Analysis of Metastatic Status, Disease Stage Dynamics, and Menopausal Influences on Breast Cancer Survival in a Cohort Study

Technical Summary: Prognostic Factors in Breast Cancer Survival

Problem Statement
Breast cancer remains the most prevalent cancer among women globally, with significant variability in occurrence, fatality, and recovery rates across populations. While early detection is associated with better outcomes, a substantial proportion of cases are diagnosed at later stages, and approximately 30% of women initially diagnosed with early-stage disease eventually develop metastatic breast cancer (MBC). Metastatic status, defined as the spread of cancer to distant sites, is a critical determinant of prognosis, with stage 4 disease carrying a 5-year survival rate of approximately 31%. Furthermore, menopausal status influences disease biology and treatment response, as postmenopausal women are more prone to hormone receptor-positive tumors. Despite advances in treatment, survival rates vary considerably, necessitating a comprehensive analysis of how metastatic patterns, disease stage dynamics, and menopausal status interact to influence overall survival (OS).

Methodology
This retrospective cohort study utilized a secondary dataset of 1,746 breast cancer cases derived from the 2018 Memorial Sloan Kettering (MSK) Cancer Center cohort, accessed via the cBioPortal for Cancer Genomics. The study population was restricted to female patients after excluding 10 male cases from an initial pool of 1,918 samples.

• Variables:
  • Independent Variables: Metastatic status (presence/absence), disease stage at diagnosis (categorized via TNM staging from IA to IV), and menopausal status (pre-, peri-, and post-menopausal).
  • Dependent Variable: Overall Survival (OS), defined as the time from diagnosis to death from any cause. Censoring was applied to patients alive at the study's end or lost to follow-up.
  • Covariates: Age, hormone receptor (HR) status, HER2 status, and overall receptor subtypes (HR+/HER2-, HR-/HER2+, etc.).
• Statistical Analysis:
  • Descriptive statistics summarized demographic and clinical characteristics.
  • Kaplan-Meier (KM) Analysis: Used to estimate survival distributions and compare survival curves across groups using log-rank tests.
  • Cox Proportional Hazards (PH) Regression: Employed to assess the association between prognostic factors and survival outcomes. The model was adjusted for potential confounders, including age and overall receptor status.
  • Assumptions: The proportional hazards assumption was verified using log-minus-log survival plots, confirming no violations. Multicollinearity was assessed via Variance Inflation Factor (VIF), with all variables falling below the threshold of 3.
  • Software: All analyses were conducted using SAS 9.4.

Key Results
The study cohort had a mean age of 52.25 years (±12.07), with the largest age group being 50–59 years (29.4%). The majority of participants were diagnosed at Stage IA (28.2%) or Stage IV (20.4%). Approximately 46.3% were pre-menopausal, 48.0% post-menopausal, and 71.8% presented with metastatic disease. The overall 5-year survival rate was 80%, with a mean survival time of approximately 15.7 years and a median of 16.5 years.

• Univariate Analysis (Log-Rank Test):

  • Cancer Stage: Statistically significant difference in survival time across stages (p < 0.001).
  • Menopausal Status: Statistically significant difference in survival time (p < 0.001).
  • Metastatic Status: The log-rank test did not show a statistically significant difference in the univariate comparison (p = 0.580), though KM plots indicated better survival for non-metastatic patients.

• Multivariate Analysis (Cox Proportional Hazards Model):
  After adjusting for age and receptor status, the following associations were identified:

  • Cancer Stage: A highly significant association with OS (p < 0.001). Stage IV exhibited the highest Hazard Ratio (HR = 4.22; 95% CI: 2.97–6.00), followed by Stage IIIB (HR = 2.90).
  • Metastatic Status: Not significantly associated with OS in the adjusted model (p = 0.448). Metastatic patients had an HR of 1.48 compared to non-metastatic patients, but this did not reach statistical significance.
  • Menopausal Status: Post-menopausal status was significantly associated with higher risk compared to pre-menopausal status (HR = 1.49; p = 0.023).
  • Receptor Status: The HR+/HER2+ subtype showed the lowest risk (HR = 0.11) compared to the reference category (HR-/HER2-), which was associated with the poorest survival outcomes.

Key Contributions
This study contributes to the literature by analyzing a large cohort (n=1,746) to simultaneously evaluate the interplay of metastatic patterns, disease stage dynamics, and menopausal influences on breast cancer survival. Key contributions include:

1. Quantification of Risk: Providing specific Hazard Ratios for advanced stages (particularly Stage IV and IIIB) and clarifying the non-significant adjusted association of metastatic status within a US-based cohort.
2. Menopausal Influence: Confirming that post-menopausal status is an independent predictor of poorer survival outcomes compared to pre-menopausal status after adjusting for covariates.
3. Methodological Rigor: Utilizing a comprehensive statistical approach that includes PH assumption verification and adjustment for key tumor characteristics (HR/HER2 status) to isolate the effects of stage and metastasis.

Significance and Claims
The authors conclude that cancer stage and menopausal status are significant determinants of survival rates in breast cancer patients, while metastatic status did not show a significant independent association after adjustment in this specific model. The findings underscore the critical importance of considering these specific factors when evaluating prognosis and directing therapeutic choices. The study highlights that while early detection remains vital, the presence of advanced staging (particularly Stage IV) drastically increases mortality risk. Additionally, the identification of post-menopausal status as a risk factor suggests that hormonal changes accompanying menopause may influence disease progression and treatment response.

The paper positions these findings as a basis for refining clinical decision-making but maintains a modest tone regarding limitations. The authors acknowledge that the study is based on data from a single institution (MSK), which may limit generalizability, and that the retrospective nature of the data introduces potential selection and information biases. Furthermore, the study assessed the number of treatments rather than specific regimens, and unmeasured confounders such as race, ethnicity, and income were not included. Consequently, the authors recommend that future research validate these findings in diverse populations and incorporate detailed treatment regimen data to deepen the understanding of survival outcomes.