Screening mammography may be less accurate in women with previous breast cancer

Screening mammography may be less accurate in women who have had breast cancer than in those who have not, according to a new study.

Researchers matched a cohort of women who had had early-stage breast cancer with women who had not. All had undergone screening mammography between 1996 and 2007 at Breast Cancer Surveillance Consortium-affiliated facilities. Women were matched by age, breast density, year of mammography and Breast Cancer Surveillance Consortium registry. Early-stage breast cancer was defined as in situ or stage I-II invasive disease. The study's primary outcomes were accuracy of mammography according to final assessment, cancer detection and interval cancer rates, and cancer stage at diagnosis. The study results appear in the Feb. 23 Journal of the American Medical Association.

Overall, 58,870 screening mammograms were performed in 19,078 women with a history of early breast cancer versus 58,870 screening mammograms in 53,315 women without. Within a year after screening, 655 cases of cancer were found in women with a personal history, 499 of which were invasive and 156 of which were in situ. Among women with no personal history, 342 cases of cancer, 285 invasive and 57 in situ, were found. Women with a history of breast cancer had cancer rates of 10.5 per 1,000 screens (95% CI, 9.7 to 11.3) versus 5.8 per 1,000 screens (95% CI, 5.2 to 6.4), a cancer detection rate of 6.8 per 1,000 screens (95% CI, 6.2 to 7.5) versus 4.4 per 1,000 screens (95% CI, 3.9 to 5.0), and an interval cancer rate of 3.6 per 1,000 screens (95% CI, 3.2 to 4.1) versus 1.4 per 1,000 screens (95% CI, 1.1 to 1.7) compared with those who had no history. Mammogram sensitivity was 65.4% (95% CI, 61.5% to 69.0%) versus 76.5% (95% CI, 71.7% to 80.7%), respectively, and specificity was 98.3% (95% CI, 98.2% to 98.4%) versus 99.0% (95% CI, 98.9% to 99.1%). Overall, 2.3% (95% CI, 2.2% to 2.5%) of mammogram results were abnormal in those with a family history versus 1.4% (95% CI, 1.3% to 1.5%) in those without. The P value for all comparisons was less than 0.001.

In women with a personal history of breast cancer, mammography screening was more sensitive for in situ disease (78.7%; 95% CI, 71.4% to 84.5%) than for invasive disease (61.1%; 95% CI, 56.6% to 65.4%) (P<0.001). Mammography was less sensitive in the first five years after a woman's first cancer than after five years had passed (60.2%; 95% CI, 54.7% to 65.5% vs. 70.8%; 95% CI, 65.4% to 75.6%) (P=0.006); sensitivity was similar for ipsilateral cancer (66.3%; 95% CI, 60.3% to 71.8%) and contralateral cancer (66.1%; 95% CI, 60.9% to 70.9%) (P=0.96). Most interval cancers and cancers detected by screening were early stage in all women regardless of personal breast cancer history.

The authors concluded that screening mammography helps detect second cases of breast cancer at an early stage in women with a personal history of early-stage disease. However, its sensitivity is lower and there is a higher rate of interval cancer than in women who have not had breast cancer. The authors noted that the higher interval cancer rates in this group may in part reflect a higher awareness and willingness to seek prompt treatment, a higher rate of adjunct screening between scheduled screening mammography exams, or a predisposition to cancers that aren't as likely to be detected with mammography. Their findings, they wrote, indicate that women with a personal history of breast cancer have “heterogeneous risk for developing another breast cancer; thus, consideration of a more tailored screening approach might be warranted” in some patients in this group.