Focus on Preventing Invasive Recurrence in Women With DCIS Does Not Sufficiently Address Breast Cancer Mortality

Get Permission

A study published recently by Giannakeas et al looked at the risk of death from breast cancer for women diagnosed with ductal carcinoma in situ (DCIS).1 The investigators anticipated that treatment would eliminate the risk of invasive ipsilateral recurrence and prevent subsequent mortality from breast cancer. However, that is not what the study found. In the study population—144,524 women with DCIS, all treated with surgery and almost half also receiving radiotherapy—“the risk of dying of breast cancer was increased threefold after a diagnosis of DCIS,” the researchers reported. “This suggests that our current treatment focus on preventing invasive recurrence is insufficient to eliminate all deaths from breast cancer after DCIS.”

To explore the study’s implications, The ASCO Post spoke with the corresponding author, Steven A. Narod, MD, Canada Research Chair in Breast Cancer, Women’s College Research Institute, and Professor, University of Toronto. “It is important for physicians and patients to be aware in their discussions that there are two different goals,” he said. “One is to prevent invasive local recurrence, and the other, to prevent death from breast cancer. Both are laudable goals, and both are associated with different strategies.” 

Steven A. Narod, MD

Steven A. Narod, MD

Chemotherapy and Tamoxifen

The risk of dying of breast cancer within 20 years of being diagnosed with DCIS was 3.4%, too low to generally recommend chemotherapy. However, for women younger than age 40 and/or Black women, “the mortality rate approached 10%, and at this level, chemotherapy might be considered,” the researchers wrote.

“The great majority of patients with DCIS don’t warrant chemotherapy,” Dr. Narod explained. “Chemotherapy might be considered for the rare patient who has two or more risk factors,” he added, but that should be decided on a patient-by-patient basis.

“I would consider tamoxifen, although I wish I had more data on tamoxifen and death from breast cancer. We don’t really know about tamoxifen, even though many people are treated with it,” he stated, commenting that the Surveillance, Epidemiology, and End Results (SEER) database does not include information on how much tamoxifen patients are using.

Some biologics also have been tried in this setting, but they have not worked well and are very expensive, Dr. Narod noted. “The idea of exploring the realms of breast cancer therapy to extend beneficial treatment to women with low risks of dying would be wonderful.”

All Had Surgery

Using the SEER registries database, researchers identified 144,524 women diagnosed with first primary DCIS between 1995 and 2014. The mean age at diagnosis was 57.4 years. All the women had surgery, and 47.1% also received radiotherapy. Women without cancer in the general population were analyzed as controls.

“DCIS itself is a cancer with metastatic potential and not a precursor for an invasive cancer with metastatic potential.”
— Steven A. Narod, MD

Tweet this quote

“Based on national incidence and case-fatality rates, the expected number of deaths from breast cancer was 458,” according to the study report, but there were actually 1,540 deaths from breast cancer. “Of them, 703 (45.7%) experienced an ipsilateral invasive recurrence or contralateral invasive breast cancer in the interval between DCIS and death from breast cancer. Among women who died, 428 (27.8%) were known to have undergone a mastectomy,” the researchers noted.

According to the study report, “The standardized mortality ratio (SMR) for death from breast cancer among women with DCIS was 3.36 (95% confidence interval [CI] = 3.20–3.53). The elevated risk of death persisted more than 15 years after diagnosis.”

“We only characterized invasive recurrence and death. We didn’t look at the nature of distant metastases,” but the deaths would have been from metastatic disease of the lung, liver, brain, and bones, Dr. Narod said.

Among women not treated with radiotherapy, the SMR was 4.12 for those who had a unilateral mastectomy and 4.14 for those who had a bilateral mastectomy. Among women treated with lumpectomy, the SMR was 2.81 for those who also received radiotherapy and 3.42 for those who did not.

Young and Black Women

“The rate of breast cancer death was nearly 12-fold higher among women diagnosed with DCIS before age 40 and 7-fold higher in Black women diagnosed with DCIS compared with the general population,” the study found. The SMR for women younger than age 40 was 11.95 vs 4.15 for women between the ages of 40 and 49 years, 2.82 for women aged 50 to 59 years, 2.65 for women aged 60 to 69 years, and 3.72 for women aged 70 to 79. Black women had an SMR of 7.56, vs 3.03 for White women, 2.40 for Southeast Asian women, and 1.89 for East Asian women.

The investigators reported that for women who were younger than 40 and/or Black, “the mortality rate approached 10%, and at this level, chemotherapy might be considered.” However, Dr. Narod noted that the 10% figure was an extrapolation, and he and his fellow researchers did not calculate rates for women who were both young and Black. “One could make the argument for chemotherapy among these women, although there is not enough evidence to to justify a consensus statement that young Black women should have chemotherapy if they are diagnosed with DCIS,” he said.

DCIS Is Breast Cancer

“DCIS itself is a cancer with metastatic potential and not a precursor for an invasive cancer with metastatic potential,” Dr. Narod stated. “If DCIS were merely a precursor of an invasive breast cancer, then preventing the invasive breast cancer should prevent the subsequent mortality,” he said.

“There is no difference between a small DCIS and a small invasive breast cancer except quantitatively: DCIS has a 3% mortality and node-negative small invasive breast cancer has a 5% mortality,” Dr. Narod noted.

“Back in 1983, Bernard Fisher showed that lumpectomy plus radiotherapy prevented local recurrence, but preventing local recurrence didn’t prevent the deaths,” he ­continued. “If you look at the SEER data for women with small breast 

cancers, there are more deaths from breast cancer than there are local invasive recurrences. Most women who die of a 1-cm breast cancer, do not experience an invasive local recurrence. Most people who die of DCIS do not experience an invasive local recurrence. From this, I’m implying that the two are similar. They vary quantitatively, but qualitatively they are part of the same spectrum.”

Different Paradigm

Patients with DCIS have a 15% chance of invasive local recurrence, Dr. Narod noted, but “preventing the invasive local recurrence has nothing to do with preventing death. It is a completely different paradigm. If it were the other way around—that DCIS is a precursor that leads to invasive breast cancer and the invasive breast cancer is the cause of the death—how is it that preventing the invasive cancer has no impact on the death?” he asked.

“We need to accept that DCIS has a metastatic potential and that studies should be done with death from breast cancer as the endpoint.”
— Steven A. Narod, MD

Tweet this quote

“We know that we prevent far more breast cancers with a mastectomy than with a lumpectomy. No one would dispute that. You have two patients with DCIS: one has her breast removed, and one merely has the DCIS removed. The one with the mastectomy has a far lower chance of invasive recurrence,” Dr. Narod said. “But the chance of dying is less for the woman with lumpectomy plus radiotherapy.”

Noninvasive Does Not Imply Nonmetastatic

Although DCIS is considered noninvasive, “that doesn’t mean nonmetastatic. The assumption is that if it doesn’t have an invasive component, it can’t metastasize. There are three ways I can question that,” Dr. Narod said.

“The simplest one is suggested by what you see on a particular slide taken at a particular point of time on a particular cross section of a DCIS. It is possible that if you looked elsewhere in place or time, that you would see an invasive component. It is possible that little tumor emboli break through the basement membrane, and then they are disseminated (and go unseen). Or it could be merely that we missed it,” he said.

“Clearly, the more invasive foci there are, the greater the chance of picking it up through a histologic cross section. DCIS has a lot of potential for invasion, but we don’t necessarily see it. To say that a single pathologic cross section at a particular point in space and time is an adequate formal representation of a dynamic process of early cancer progression is wishful thinking. Maybe microinvasion is a continuous thing,” Dr. Narod suggested.

“The second hypothesis is that maybe the only way of dissemination is not lymphatic or blood borne, but there are mechanisms, not yet known, which allow single cancer cells or a small bolus of cancer cells to escape an acinus or a ductal lobular unit,” he said.

The third hypothesis, which “leads to the most speculation, is that both DCIS and invasive cancers are manifestations of something that happened in an earlier phase, when there is transient rapid dissemination of cancer cells throughout the body from somewhere in the breast, unbeknownst to us, and that later they manifest as intraductal cancers; intrabreast cancers; nodal cancers; or cancer in the lung, liver, brain, or bones. In other words, there is a common precursor to all of them, and we are seeing a later manifestation of something that happened at a microlevel years in the past,” he said.

Death as the Endpoint

Most people would acknowledge that DCIS is heterogeneous, and some women have a higher chance of developing breast cancer. “If we knew who they were, we could give the more aggressive treatment to those people. This is the universal sentiment for personalized medicine,” Dr. Narod said. “The problem with that concept is that it pertains to the prevention of invasive recurrence, and the risk factors for invasive recurrence and the risk factors for death are not necessarily the same.”

Thus, to save more lives of women diagnosed with DCIS requires looking at factors beyond those predicting invasive local recurrence. “We need to accept that DCIS has a metastatic potential and that studies should be done with death from breast cancer as the endpoint,” Dr. Narod said.

“Studies like ours—and there are three or four—are observational studies that look at the risk of breast cancer death after a diagnosis of DCIS, and they all showed the same thing,” Dr. Narod said. “The risk is there, but it is really low, and it doesn’t seem to vary much whether or not you give any treatment. The treatments are designed to prevent invasive local recurrence.” 

DISCLOSURE: Dr. Narod reported no conflicts of interest.


1. Giannakeas V, Sopik V, Narod SA: Association of a diagnosis of ductal carcinoma in situ with death from breast cancer. JAMA Netw Open 3:e2017124, 2020.


Related Articles

Expect Questions About Increased Cancer Mortality Risk Among Patients Diagnosed With DCIS

A large cohort study1 finding that the risk of dying of breast cancer was increased threefold after a DCIS diagnosis may cause patients diagnosed with DCIS to ask what they can do to reduce that risk. Currently, there is little that most patients can do. “The lifetime risk of death following DCIS...