Early detection of breast cancer the second time around: mammography in women with a personal history of breast cancer

Women with a personal history of breast cancer (PHBC) represent an increasing group of cancer survivors, and have a lifelong increased risk of developing a new or recurrent cancer in the conserved (ipsilateral) breast, or a contralateral cancer. The risk of a “second” breast cancer in women with PHBC has been estimated at 5.4 to 6.6/1000 woman-years.1 Evidence of screening benefit in PHBC women comes from observational studies2-4 and extrapolation of benefit from randomised mammographic screening trials; consensus-based recommendations include annual mammography in routine surveillance of PHBC women.5-7 Early detection may also minimise the physical and psychosocial burden and consequences of a second breast cancer. Evidence reviews have consistently acknowledged the lack of quality data on mammographic screening in PHBC women,4,8 and research into screening high-risk women has mostly focused on those with breast and ovarian cancer susceptibility genes; evaluation of screening in women with PHBC has received relatively little attention.4,8

Together with other collaborators from Australia and the United States, we recently reported the most comprehensive study to date of mammographic screening in women with a history of early-stage breast cancer.9 Based on data from mammography facilities affiliated with the US National Cancer Institute-funded Breast Cancer Surveillance Consortium for the years 1996–2007, we examined the accuracy and outcomes of screening in 58 870 mammograms from 19 078 women with PHBC. These were matched to 58 870 screening mammograms from 55 315 women without PHBC, to ensure that characteristics that affect mammography accuracy (such as age group and breast density) were similar in both cohorts.9

Key findings from the study are summarised in the Box. Although the cancer detection rate of screening was substantially higher in PHBC women relative to women without PHBC (6.8 v 4.4/1000 screens), screening detected a smaller proportion of the cancers occurring in PHBC women at 1-year follow-up (65.4% v 76.5%). Due to this lower sensitivity of mammographic screening, as well as the higher underlying breast cancer risk in PHBC women, the interval cancer rate in women with PHBC was more than twice the rate in women without PHBC (3.6 v 1.4/1000 screens). The lower sensitivity and higher interval cancer rate occurred despite more imaging work-up in PHBC women relative to women without PHBC (18.1% v 8.3%) and more recommendations for biopsy or surgical consultation (2.3% v 1.4%).

Unique aspects of this study were characterisation of screening outcomes and underlying cancer risk according to characteristics of the women and their first cancer treatment, and ascertainment of ipsilateral and contralateral second cancers.9 The latter has not previously been possible because notification of recurrent cancer is not obligatory in most countries, but was achieved through the data-linkage processes of the Breast Cancer Surveillance Consortium. Hence, the twofold higher rate of breast cancer in women with PHBC relative to those without (Box) represents a reliable risk estimate. We found heterogeneity of both breast cancer risk and screening accuracy across subgroups of women and by the treatment received for the first cancer. The highest rates of second cancers (> 12 cancers/1000 screens) were in women younger than 50 years, women with extremely dense breasts, women who received breast-conserving surgery without radiation therapy or who did not receive any systemic therapy, and women with previous ductal carcinoma-in-situ (the latter probably reflects the infrequent use of adjuvant therapy in the treatment of ductal carcinoma-in-situ).

Should these findings change our practice or recommendations for women with PHBC? Superficially, these results9 could be interpreted as casting doubt on the efficacy of mammographic screening in women with PHBC. However, careful consideration of the data shows that although mammographic screening was relatively less sensitive in PHBC women, it detected their second cancers at an early stage, with similar stage distributions to women without PHBC.9 The evidence from this study therefore supports consensus recommendations that advise mammographic screening in women with PHBC, and which also include clinical breast examination as part of annual surveillance.5-7 Many of the interval cancers were also early-stage, suggesting that cancers not detected at time of mammographic screening were detected through clinical examination or investigations prompted by symptoms, or possibly through ad-hoc adjunct screening (using breast ultrasound or magnetic resonance imaging). Currently, adjunct screening is not routinely recommended for women with PHBC (unless they are known to have breast cancer gene mutations), and our study did not evaluate adjunct imaging. Our data may, however, guide future evaluations of adjunct screening in subgroups of PHBC women who had high interval cancer rates; specifically, women younger than 50 years and/or with extremely dense breasts, and those who had breast conservation without radiation therapy for their first cancer. It is essential that evaluations of strategies integrating adjunct screening for PHBC women consider the impact this may have on false positives and overdetection, as well as whether enhanced screening sensitivity through adjunct imaging translates into a reduction in interval cancers.

Recommendations from the BreastScreen Australia evaluation10 included the potential provision of annual screening through BreastScreen for women with PHBC, from 5 years after their first cancer diagnosis. Our study shows that screening sensitivity improves after 5 years from the first cancer diagnosis and supports the idea of allowing PHBC women access to screening in BreastScreen-accredited facilities, which undergo rigorous quality assurance, to ensure that they receive high-quality mammographic screening. BreastScreen may also be ideally placed to implement screening evaluations for PHBC women to provide insight into screening outcomes in these women in the Australian setting.