How to individualize cancer risk reduction after a diagnosis of DCIS

A recent study reported in JAMA 
Oncology evaluated 10-year and 20-year breast cancer–specific mortality following diagnosis and treatment of ductal carcinoma in situ (DCIS) using the Surveillance, Epidemiology, and End Results (SEER) registries.¹ The study included cases of pure DCIS (without lobular carcinoma in situ or microinvasion) diagnosed from 1988 to 2011 among women younger than age 70. It evaluated variables including age, race, income, type of surgery, radiation, subsequent diagnoses of invasive primary breast cancer, and, when applicable, cause of death.

Overall mortality rate was 3.3%
Mean follow-up was 7.5 years, with a 20-year breast cancer–specific mortality rate of 3.3% overall. Mortality was higher among young women diagnosed before the age of 35 years (7.8% vs 3.2%), and among black women (7.0% vs 3.0% for white women). The risk of dying from breast cancer was 18 times higher for women who developed subsequent ipsilateral invasive breast cancer. Mortality also was related to adverse DCIS characteristics such as grade, size, comedo-necrosis, and lack of an estrogen receptor.

Among patients who underwent lumpectomy, the addition of radiation reduced the risk of subsequent ipsilateral invasive breast cancer at 10 years (2.5% vs 4.9%; P<.001). However, radiation did not improve the 10-year rate of breast cancer mortality (0.8% for women who had lumpectomy with radiation, 0.9% for women who had lumpectomy alone, and 1.3% for women with unilateral mastectomy).

The prevention of ipsilateral invasive recurrence with radiation did not reduce mortality rates, as more than 50% of the women who died of breast cancer did not have an ipsilateral invasive recurrence prior to their death.

How these findings fit 
into the larger picture
The findings of this landmark study confirm earlier reports, which showed that radiation after lumpectomy can reduce local recurrence but does not improve survival.²

Likewise, mastectomy, when compared with lumpectomy, offers no survival benefit and does not represent appropriate therapy for most women with small, unifocal DCIS.³

DCIS itself is not a life-threatening condition and has been described as a precursor lesion that, over 10 to 40 years, can lead to the development of invasive disease (FIGURE).^4,5 High-grade DCIS tends to lead to high-grade invasive ductal carcinoma, and low-grade DCIS may develop into low-grade invasive disease.⁶

What is DCIS?In DCIS, abnormal cells develop and multiply but remain contained wholly within the milk duct. Once these cells spread beyound the duct to breast tissue, the cancer becomes invasive.

The increasing prevalence of screening mammography means that more small in situ lesions are being identified in US women. Unlike colonoscopy, which can prevent colon cancer by removing colon polyps, mammography with subsequent surgical treatment of DCIS has not reduced the incidence of invasive breast cancer.⁷This 
finding leads us to question whether all DCIS should be considered a precursor lesion. This well publicized study is generating controversy regarding overdiagnosis and overtreatment of DCIS.

Limitations of this study
The majority of patients in the SEER registry underwent surgical treatment of DCIS with or without radiation and had a survival rate of more than 97%. Because there was no untreated control group, this study does not allow us to draw any inferences on the role of expectant management of DCIS.

Although it is often declined by patients, tamoxifen reduces the risk of ipsilateral and contralateral invasive and in situ breast cancer. Regrettably, information on the use of adjuvant hormonal therapy after an initial diagnosis of DCIS was not included in this analysis.

Why did death from invasive 
cancer sometimes follow a 
diagnosis of DCIS?
Several factors could have contributed to the 3% mortality rate from invasive breast cancer among women in this large study of DCIS. For one, it is challenging for pathologists to perform comprehensive tissue sampling of mastectomy specimens—or even large lumpectomy specimens. Accordingly, occult microinvasive disease could be missed.^8,9 As a result, occult invasive disease could go untreated, which could have contributed to the breast cancer mortality observed in this study.

Recommendations for practice
How can we better predict the behavior of DCIS and tailor treatment based on the biological behavior of each patient’s disease?

Individualize therapy. The likelihood of local invasive breast cancer recurrence should be estimated for each patient based on the size and grade of her disease. Furthermore, genetic profiling of DCIS has been developed with the Oncotype DX test (Genomic Health) multigene assay. This test can be performed on pathology specimens and has been shown to estimate the risk of in situ and invasive in-breast recurrence in patients who have undergone margin-negative lumpectomy for DCIS and who prefer to avoid radiation but are willing to take tamoxifen.¹⁰