Number of prior negative screening outcomes does not influence future risk of breast cancer

Marie Lilleborge,Ragnhild S Falk,Solveig Hofvind

doi:10.1007/s10654-020-00645-0

Abstract

We questioned whether a history of negative screening outcomes could be used to predict breast cancer risk, and thus be used as a potential factor for stratification of mammographic screening. Data from the Norwegian population based breast cancer screening program, BreastScreen Norway, was used to estimate cumulative hazard rates for breast cancer by number of prior negative screening outcomes among participants from 1995 through 2016. We followed three age cohorts of women, who started screening at age 50–54, 55–59, and 60–64 years. Further, we estimated the absolute and relative risk of breast cancer by number of prior negative screening outcomes. The cumulative hazard curves were parallel for all numbers of negative screening outcomes for all age cohorts. The absolute risk of breast cancer increased with number of negative screening outcomes for the youngest age cohort. For the oldest age cohorts, the absolute risk was stable during the screening period and decreased thereafter. The number of negative screening outcomes was not associated with risk of breast cancer, adjusted for age, percent screening attendance and calendar years (HR 1.00, 95% CI 0.98–1.02). Our results suggest that the number of negative screening outcomes does not predict breast cancer risk among participants in BreastScreen Norway.

Highlights

Mammographic screening is aimed at reducing breast cancer mortality by detecting tumors at an early stage
In age cohorts 2 and 3, more than 85% of the women had their first screen during the staggered implementation of BreastScreen Norway, 1995–2005
We did not observe any reduction in breast cancer risk among women with high versus low number of negative screening outcomes, in BreastScreen Norway

Summary

Introduction

Mammographic screening is aimed at reducing breast cancer mortality by detecting tumors at an early stage. A reduced rate of breast cancer is expected for a time after a negative screening outcome because of earlier diagnoses due to screening. Inspired by Walter and Day [5], Andersen et al introduced a hypothesis that women with a certain number of prior negative screening outcomes could be identified as a low risk group [6]. They used data from organized mammographic screening in Sweden and Denmark, and found negative screening outcomes were not a predictor of breast cancer risk, and not suitable for stratified screening. For the purpose of complementing, replicating, and potentially expanding the generalizability of the findings by Andersen et al, we wanted to investigate the same research question

Methods

Results

Conclusion