Context: Selection of systemic adjuvant therapies for women diagnosed as having breast cancer is based on risk estimations for cancer recurrence. In such estimations, tumors detected by mammography screening are considered to be associated with a similar risk of recurrence as tumors of similar size found by other methods.

Objective: To compare the risk of recurrence and survival among women with cancerous tumors detected by mammography screening compared with other methods (outside of screening).

Design, setting, and patients: Retrospective study comparing clinical, histopathological, and biological features of cancerous tumors detected by mammography screening compared with tumors detected outside of screening. Women diagnosed as having breast cancer in 1991 or 1992 were identified from the Finnish Cancer Registry (n = 2842). The median follow-up time was 9.5 years. Cancer biological variables were analyzed from tumor tissue microarrays using immunohistochemistry or in situ hybridization and included ERBB2, TP53, and MK167 expression and ERBB2 amplification data.

Main outcome measures: Univariate and multivariate analyses of potential risk factors for distant recurrence of breast cancer and 10-year survival.

Results: Of the 1983 women with unilateral invasive breast cancer, data on tumor diameter were available for 1918 women. Women with cancerous tumors detected by mammography screening had better estimated 10-year distant disease-free survival than women with tumors found outside of screening (tumor size of < or =10 mm [n = 386] 92% vs 85% [P =.04]; 11-20 mm [n = 808] 88% vs 76% [P<.001]; 21-30 mm [n = 409] 86% vs 63% [P =.008]; >30 mm [n = 315] 68% vs 50% [P =.12], respectively). In a Cox multivariate model that included cancer biological factors, the relative hazard ratio for distant recurrence among women with tumors detected outside of screening (HR, 1.90; 95% confidence interval, 1.15-3.11) was significantly higher than among women with tumors detected by mammography screening (P =.01). Breast cancer diagnosis by mammography screening was an independent prognostic variable reducing the relative HR for distant recurrence. This effect was equal to or greater than the effect of 1-cm decrease in tumor diameter (HR, 1.20; 95% confidence interval, 1.10-1.31).

Conclusions: Cancerous tumors detected by mammography screening are associated with a better prognosis than tumors of similar size found outside of screening. The risk of distant metastases is overestimated for women diagnosed as having cancer by mammography screening unless the method of detection is taken into account in risk estimations.