A new study has reported that screening mammograms designed to detect new breast cancer in breast cancer survivors are less accurate than screening mammograms in women without a personal history of breast cancer. Breast cancer survivors are at risk of developing another breast cancer. However, high-quality data on screening performance in such women is scarce. The study included 19,078 women with a history of breast cancer who were mammogram matched (breast density, age group, year of mammogram) to 55,315 women without a history of breast cancer (the controls). All of the women were screened during 1996 to 2007 in facilities affiliated with the Breast Cancer Surveillance Consortium. Included in the study were a total of 58,870 screening mammograms in the survivors (who had a history of either noninvasive (in situ) or early-stage breast cancer) and 58,870 matched screening mammograms in the controls.
Screening-detected and interval cancers (cancers found between screening visits) in women with and without a personal history of breast cancer were predominantly early stage. A total of 655 new cancers were found within one year after screening among the women with a history of breast cancer women (499 invasive, 156 in situ), whereas 342 cancers (285 invasive, 57 in situ) were found in the controls. The breast cancer survivors had 10.5 new breast cancers per 1,000 screens, however the cancer detection rate was only 6.8 per 1,000 screens. Women without a history of breast cancer had a rate of 5.8 breast cancers per 1,000 screens while their cancer detection rate was 4.4. Cancer survivors had an interval cancer rate of 3.6 per 1000 screens compared to 1.4 per 1000 screens among controls.
Mammogram sensitivity was 65.4% among survivors compared to 76.5% among controls and specificity was 98.3% in survivors compared to 99.0% in controls. Mammogram sensitivity refers its ability to detect breast cancer. A very sensitive test may result in a number of false-positive results, but almost no true cases of the disease will be missed. Mammogram specificity refers to the fraction of the time that the screen yields true negatives. The closer the specificity is to 100%, the more likely a negative mammogram means that the patient is actually free of breast cancer. Mammogram sensitivity in breast cancer survivors was higher for detection of in situ disease (78.7%) than detection of invasive breast cancer (61.1%). Mammogram sensitivity was lower during the first five years (60.2%) than after five years from first cancer (70.8%) and was similar for detection of cancer in the same breast as the original tumor (ipsilateral) (66.3%) and cancer in the opposite breast (contralateral) (66.1%). The authors conclude that mammographic screening in women with a personal history of breast cancer has lower sensitivity and a higher interval cancer rate compared to women without a personal history of breast cancer despite more evaluation.
Comments regarding the study
The study results may appear surprising since mammograms of breast cancer survivors are given extra scrutiny. Part of the reason for the reduced accuracy of screening mammograms in such women may be the higher likelihood of dense breasts (which are harder to read) among breast cancer survivors, although the authors attempted to control for this factor. The presence of scar tissue from breast cancer surgery may also play a role. The authors also found that treatment for breast cancer had an effect on the accuracy of mammograms. The accuracy was higher among women who had undergone a mastectomy than among those who had undergone breast-conserving therapy such as lumpectomy. Radiotherapy was associated with a very small reduction in accuracy and women who had undergone chemotherapy were less likely to have their cancer detected by mammography than women who did not. The study findings do not suggest that current screening guidelines be altered, according to the authors and several commentators. MRIs may be more sensitive than mammograms, but they are likely to lead to many more biopsies for what may prove to be benign findings.
Perhaps the most relevant finding for breast cancer survivors is the relatively high rate of interval cancers. These are breast cancers (typically found by the women themselves) that are diagnosed between screening mammograms. This indicates that survivors should pay attention to any changes in their breasts and bring them to the attention of their oncologists even if they have recently undergone mammography. This will not be a welcome message to many women who would like to put the anxiety associated with a breast cancer diagnosis behind them. They may also feel that, having been identified as prone to breast cancer, they of all people should be able to rely on regular screening to detect any recurrences in the breast. One solution might be to schedule one day per month to perform a thorough breast self exam and then give oneself permission not to think about the possibility of breast cancer recurrence on any other day of the month.