Factors Associated with Early Mortality Among Patients with De Novo Metastatic Breast Cancer: A Population-Based Study

Abstract

Background: Although improvements in survival have been achieved for patients with metastatic breast cancer, some patients experience early death after diagnosis.

Patients and methods: Using Surveillance, Epidemiology, and End Results data, we identified 26,538 patients with de novo metastatic breast cancer diagnosed between January 1, 2000 and June 30, 2011. We evaluated time trends for deaths at 1 and 6 months after diagnosis. We then restricted the cohort to patients diagnosed between 2010 and 2011 (n = 3,317), when human epidermal growth factor receptor 2 was routinely collected, and examined factors associated with early death.

Results: In 2000, 15.9% of patients died within 1 month of diagnosis and 33.2% within 6 months. In 2011, the proportion of women dying within 1 month decreased to 13.4% and 26.3% within 6 months (p < .001). Older age and uninsured status were associated with early death (at both time points, age ≥70 [versus age <40] had >8.5 higher odds of dying, and uninsured [versus insured] patients had >2.5 higher odds of death). In addition, in some subgroups (e.g., no insurance and triple negative disease), more than half of patients died within 6 months. Region was also associated with early death.

Conclusion: Although we observed improvements in the proportion of patients experiencing early death, one quarter of patients with de novo metastatic disease diagnosed in 2011 died within 6 months of diagnosis. In addition to tumor factors and older age, geography and uninsured status were associated with early death. Our findings highlight the need for focused interventions for metastatic patients at highest risk for poor outcomes. The Oncologist 2017;22:386-393 IMPLICATIONS FOR PRACTICE: With nearly one quarter of patients in our dataset diagnosed in 2011 dying within 6 months of diagnosis, our findings highlight the persistent and critical need of further characterization and identification of patients who are risk for poor outcomes in order to optimize care, impact change, and improve outcomes for all women with metastatic breast cancer. Our data also emphasize the need for interventions among those at highest risk for early death. These interventions would likely promote immediate referral for clinical trial participation, early palliative care referrals, and additional supportive services, optimizing equitable patient access to cancer treatment and care.

Keywords: Breast cancer subtypes; End Results (SEER); Epidemiology; Metastatic breast cancer; Mortality; Surveillance.

Figure 1.

Flow diagram of patient population. Abbreviations: HER2, Human epidermal growth factor; HR, hormone receptor.

Figure 2.

Time trends of death by 1 and 6 months. Note: Cochran‐Armitage trend test was used to test death proportions over time.

Figure 3.

Unadjusted proportions of women dying by 1 and 6 months by insurance and subtype. Uninsured includes uninsured patients and patients with unknown insurance status. Insured patients include insured patients and insured not otherwise specified. *, Cells suppressed due to sample size <11. ¶, Wilson procedure without a correction for continuity. Abbreviations: CI, confidence interval; HER2, Human epidermal growth factor; HR, hormone receptor.