Long-term crude probabilities of death among breast cancer patients by age and stage: a population-based survival study in Northeastern Spain (Girona-Tarragona 1985-2004)

Abstract

Background: We provide population-based long-term survival indicators of breast cancer patients by quantifying the observed survival, and the probabilities of death due to breast cancer and to other causes by age and tumor stage at diagnosis. Methods: We included a total of 10,195 female patients diagnosed before 85 years with invasive primary breast cancer in Girona and Tarragona during the periods 1985-1994 and 1995-2004 and followed-up until December 31st 2014. The survival indicators were estimated at 5, 10, 15 and 20 years of follow-up comparing diagnostic periods. Results: Comparing diagnostic periods: I) the probability of death due to other causes did not change; II) the 20-year survival for women diagnosed ≤ 49 years increased 13% (1995-2004 = 68%; 1985-1994:55%), whereas their probability of death due to breast cancer decreased at the same pace (1995-2004 = 29%; 1985-1994 = 42%); III) at 10 years of follow-up, decreases in the probabilities of death due to breast cancer across age groups switched from 11 to 17% resulting in a risk of death reduction of 19% after adjusting by stage. During 1995-2004, the stage-specific 10-year probabilities of death due to breast cancer switched from: 3-6% in stage I, 18-20% in stage II, 34-46% in stage III and surpassed 70% in stage IV beyond 5 years after diagnosis. Conclusions: In our study, women diagnosed with breast cancer had higher long-term probability to die from breast cancer than from other causes. The improvements in treatment and the lead-time bias in detecting cancer in an early stage resulted in a reduction of 19% in the risk of death between diagnostic periods.