Risk-dependent conditional survival analysis and annual hazard rate of inflammatory breast cancer

Abstract

Purpose: The real-time prognosis of patients with inflammatory breast cancer (IBC) after surviving for several years was unclear. We aimed to estimate survival over time in IBC using conditional survival (CS) and annual hazard functions.

Patients and methods: This study recruited 679 patients diagnosed with IBC between 2010 and 2019 from the Surveillance, Epidemiology, and End Results (SEER) database. We used the Kaplan-Meier method to estimate overall survival (OS). CS was the probability of surviving for another y years after surviving for x years after the diagnosis, and the annual hazard rate was the cumulative mortality rate of follow-up patients. Cox regression analyses were used to identify prognostic factors, and changes in real-time survival and immediate mortality in surviving patients were assessed within these prognostic factors.

Results: CS analysis showed real-time improvement in survival, with 5-year OS updated annually from the initial 43.5% to 52.2%, 65.3%, 78.5%, and 89.0% (surviving 1-4 years, respectively). However, this improvement was relatively small in the first two years after diagnosis, and the smoothed annual hazard rate curve showed increasing mortality during this period. Cox regression identified seven unfavorable factors at diagnosis, but only distant metastases remained after five years of survival. Analysis of the annual hazard rate curves showed that mortality continued to decrease for most survivors, except for metastatic IBC.

Conclusion: Real-time survival of IBC improved dynamically over time, and the magnitude of this improvement was non-linear, depending on survival time and clinicopathological characteristics.

Keywords: Annual hazard rate; Conditional survival; Inflammatory breast cancer; Overall survival; Prognostic factor.