Research Brief: Breast Cancer Mortality After a Diagnosis of DCIS

Research Briefs BoxA recent study by Dr. Steven A. Narod and colleagues provides evidence that for most women diagnosed with ductal carcinoma in situ, DCIS, the condition will not impact life expectancy. It supports a mounting call for research to determine which cases of DCIS would benefit from milder interventions or preventive strategies rather than more aggressive treatments. But for the very small percentage of uncommon DCIS cases that ultimately prove lethal (less than one percent), research is needed to find more effective treatments.


DCIS (ductal carcinoma in situ) is frequently called a “stage 0” cancer, meaning that this collection of abnormal cells is noninvasive, or has not spread from the milk duct to other parts of the breast. Before mammography screening became widespread, DCIS was rare. It accounted for about 3 percent of new breast cancer cases and was usually detected by noticeable symptoms such as being large enough to feel or accompanying bloody nipple discharge. Today, DCIS is detected with mammograms, often from its association with microcalcifications (tiny deposits of calcium salts) that are easily spotted with imaging technology. The number of DCIS cases diagnosed each year through mass screening programs now reaches up to 60 thousand American women and accounts for 25 percent of new breast cancer cases.

For decades, the prevailing theory was that DCIS functioned as the first stage in a progression toward life-threatening disease. In turn, treating DCIS like an invasive, also called infiltrating, breast cancer (one that has spread beyond where it developed and is growing into surrounding, healthy tissues) would decrease the prevalence of advanced breast cancers and consequently reduce mortality. This typically meant doing surgeries such as mastectomy or lumpectomy with or without radiation. Unfortunately, treating DCIS as a cancer in waiting did not result in a decline in advanced breast cancer or the number of deaths from the disease. [1]

This scenario raises questions about what constitutes appropriate treatment and what may in the long run be considered overtreatment (treatment of a cancer that would have gone away on its own or never caused symptoms.) [2] Enduring controversy about whether or not DCIS should remain classified as an early version of breast cancer remains.

The Latest Study

A recent study published in JAMA Oncology (August 20, 2015), “Breast Cancer Mortality After a Diagnosis of Ductal Carcinoma In Situ,” analyzed the largest collection of data on DCIS examined to date. [3] Dr. Steven A. Narod, M.D., of the Women’s College Research Institute in Toronto and his colleagues used records from population-based cancer registries managed by the U.S. National Cancer Institute, the Surveillance, Epidemiology, and End Results Program, or SEER, to analyze two decades of data (1998 to 2011) on more than 100,000 women diagnosed with DCIS before age 70. Nearly every woman was treated for DCIS with mastectomy or lumpectomy, with or without radiation. Since DCIS has a very high survival rate, large numbers of DCIS cases were needed to assess whether treatment affected mortality and to identify high-risk factors for premature death. (Note: SEER data document 28 percent of cancer cases in the U.S.)

To compare the mortality risk of the women in the analysis to all U.S. women, results were adjusted using the Kaplan-Meier Method (a statistical process for estimating survival over time) to construct a Standardized Mortality Ratio. The average risk of dying from DCIS, calculated to be 3.3 percent at 20 years, was 1.8 times the average lifetime risk of comparably aged women dying from invasive breast cancer in their lifetime. For women over age 65, the increase was 1.4 times that risk. With the exception of women in high-risk categories, the data suggest that the risk of death due to DCIS does not increase the average lifetime risk of death from invasive breast cancer. Less than 1 percent of the women died from breast cancer (other studies range as high as 5 percent), suggesting that most cases of DCIS (95 to 99 percent) are not precursors to invasive cancer.

For U.S. women with DCIS, the SEER data yielded three key factors associated with a higher risk of dying from breast cancer. The deaths were concentrated among black women and women diagnosed with DCIS at a young age (under age 40, especially under 35). The deadly cases were far more likely to have high-risk biomarkers, with tissue lacking estrogen and progesterone receptors (ER/PR-negative) or producing a protein called human epidermal growth factor receptor 2 (HER2), a biomarker that represents poor prognoses in invasive breast cancer as well. More aggressive treatment did not improve survival for these high-risk groups. For most DCIS diagnoses, which were low risk, treatment did nothing to change the overall outcome. Women treated for DCIS had about the same chances of dying of breast cancer as women in the general population.


An editorial by Laura Esserman, M.D., M.B.A., and Christina Yau, Ph.D., of the University of California, San Francisco, on “Rethinking the Standard for Ductal Carcinoma In Situ Treatment” accompanied the study’s publication to extend the interpretation of study results. [4] They argue:

First, there was a very low risk of dying from DCIS — less than 1 percent in the 20 years of SEER data, compared with 5 percent who died from other causes during that same period. For every woman who died from breast cancer after DCIS, five died from causes other than breast cancer, such as heart failure or lung cancer. Thus, women with DCIS who are not in high-risk groups and at very low mortality risk may be amenable to a “watch-and-wait” protocol or to alternative non-invasive treatments.

Second, the exceptional cases with considerably higher mortality risk suggest that there are different types of DCIS. For example, since women under age 35 to 40 rarely get mammogram screens, the DCIS in this age group would more likely be symptomatic (with a palpable mass or bloody nipple discharge), distinct from screen-identified cases, and women under age 40 made up 5 percent of all cases in the study analysis. Esserman and Yau estimate from study data that 80 percent of women lack the high-risk factors identified associated with increased mortality (young age, black ethnicity, and high-risk biomarkers), suggesting that most women with DCIS are at about the same risk of dying from breast cancer as women in the general population. [5]

Third, more aggressive treatments such as mastectomy versus lumpectomy, and radiation after lumpectomy did not improve survival. Reserving that protocol for high-risk cases and looking for alternative treatments (such as no external breast radiation) for most DCIS cases may be the prudent choice. Esserman and Yau call for new treatment strategies for both low-risk and high-risk DCIS.

Finally, while Narod and his colleagues emphasize that there are similarities between certain DCIS cases and invasive disease (such as higher risk of death for cases with particular biomarkers (e.g., ER/PR- and/or HER2+), the data suggest that having a recurrence of a second DCIS in the same breast or one in the other breast did not increase risk of death; only an invasive recurrence did. Primary DCIS and recurrences acted instead like atypical cell growths (called ‘benign’) that are risk factors for breast cancer, but not precursors to cancer. Esserman and Yau stress that similarities between DCIS and invasive breast cancer would likely exist only within the very high-risk DCIS cases.

Most women (over 80 percent) with DCIS had none of the major factors associated with high risk of premature death (being under age 40, African heritage, ER/PR-, or HER2+). The other 20 percent had one or more of these risk factors; still, less than 1 percent of cases overall died from breast cancer. When DCIS was diagnosed in women under age 35, the risk of dying from breast cancer was 17 times greater than expected for women of comparable age, making very young age at diagnosis a crucial risk factor. Distinguishing the different types of DCIS should inform new standards for treating low- and high-risk groups to maximize benefits and minimize the harms.

[1] H. Gilbert Welch and William C. Black, “Overdiagnosis in Cancer,” J Natl Cancer Inst 2010;102:605-613.

[2] Laura Esserman et al. “Addressing overdiagnosis and overtreatments in cancer: a prescription for change.” Lancet Oncol.2014;15(6):e234-e242.

[3] Steven A. Narod et al., “Breast Cancer Mortality After a Diagnosis of Ductal Carcinoma In Situ,” JAMA Oncology, Aug.20, 2015, online. Doi:10.1001/jamaoncol.2015.2510.

[4] Laura Esserman and Christina Yau, “Rethinking the Standard for Ductal Carcinoma In Situ Treatment,” JAMA Oncol. Aug.20, 2015, online. Doi:10.1001/jamaoncol.2015.2607.

[5] ACS Facts & Figures, 2015. Narod et al. converted the SEER data into Standardized Mortality Ratios, adjusted rates comparing a subgroup with all women, but sometimes broken down into age groups. Thus, the overall 1 percent mortality after 20 years became 1.1 percent SMR after 10 years and 3.3 percent SMR after 20 years.

Be Sociable, Share!

Articles & Posts