The relationship between DCIS Score risk group and 10-year risk of an ipsilateral breast event, whether in situ or invasive, was highly statistically significant (P = 0.006 for any ipsilateral breast event and P = 0.003 for an invasive ipsilateral breast event) [18]. This was true with or without the adjustment for tamoxifen use. For the 230 patients with a low-risk DCIS Score, the 10-year risk of any ipsilateral breast event was 10.6 % (95 % CI 6.9–16.2 %) and 3.7 % (95 % CI 1.8–7.7 %) for an invasive breast event. For the 53 patients with an intermediate-risk DCIS Score, the 10-year risk of any ipsilateral breast event was 26.7 % (95 % CI 16.2–41.9 %) and 12.3 % (95 % CI 5.1–27.8 %) for an invasive breast event. For the 44 patients with a high-risk DCIS Score, the 10-year risk of any ipsilateral breast event was 25.9 % (14.8–43.1 %) and 19.2 % (95 % CI 9.5–36.4 %) for an invasive breast event. Multivariable models of risk for ipsilateral breast events, both excluding and including the DCIS Score, were also performed. When the DCIS Score was excluded, tumor size and menopausal status were the only factors significantly associated with risk for an ipsilateral breast event (HR 1.54, P = 0.006 and HR 0.49, P = 0.02, respectively). However, when the DCIS Score was included in the model, the 12-gene score was statistically significant (HR 2.37, P = 0.02), in addition to tumor size and menopausal status. Tumor grade and the presence of comedonecrosis were not associated with a risk for ipsilateral breast events.

There is currently no known subset of patients with DCIS who do not benefit from adjuvant radiation for risk reduction following breast-conserving therapy. This is true regardless of patient’s age, tumor size, tumor grade, and margin status [30]. In the Early Breast Cancer Trialists’ Collaborative Group analysis, the risk of an ipsilateral breast event was decreased from 18.1 to 7.6 % at 5 years and 28.1–12.9 % at 10 years in patients with ≤ 2 cm of DCIS who underwent adjuvant radiation [30]. Despite this, survival is equivalent for excision alone and excision followed by adjuvant radiation therapy. Radiation therapy is not without risk. There are short-term side effects and potential long-term consequences, some of which can be devastating. In addition, the therapy is quite costly and access to care can be difficult for many patients. Are these risks and costs justified for all patients with DCIS?

The ability to identify a subset of patients who could safely omit radiation is desirable. Thus, patients who have undergone wide local excision for DCIS who have a low-risk score could reasonably elect to omit radiation therapy, while patients in the intermediate or high-risk groups should consider adjuvant radiation therapy and tamoxifen following wide local excision.

Despite these promising findings, the DCIS Score has not yet been universally adopted for making treatment decisions. There are several important limitations. First, the E5194 study used to validate the DCIS Score included a narrow subset of patients with DCIS. Specifically, the study included patients already considered to be at low risk for recurrence, and included patients with small volumes of DCIS and with margins of at least 3 mm. Therefore, these results cannot be applied to women with larger volumes of DCIS or with closer surgical margins. In addition, only 30 % of the patients in the E5194 study received adjuvant tamoxifen therapy, and the absolute number of tamoxifen-treated patients included in the validation study was only 96 [18]. Because tamoxifen decreases the risk of both local recurrence and contralateral breast cancer events in patients with ER-positive DCIS [6, 31], use of tamoxifen would likely mean fewer recurrences for patients with all DCIS Scores. The ER gene was specifically excluded from the DCIS Score given that it specifically predicts benefit with tamoxifen therapy. For patients with ER-negative DCIS, the DCIS Score has not been adequately studied; only nine patients with ER-negative DCIS were included in the validation study (less than 3 % of the total cohort). Additional study is needed to verify these validation results across broader populations of patients with DCIS.

Although the DCIS Score is promising for stratifying patients by risk of in-breast recurrence, it does not supplant other patient and tumor characteristics such as tumor size and menopausal status, which remained significant predictors of recurrence with and without inclusion of the DCIS Score on multivariate analysis, P < 0.02 for each [18]. The DCIS Score is therefore a helpful addition to, but not replacement for, the use of already available clinicopathologic features used to make treatment decisions.

It is important to note that the DCIS Score does not equate with recommendations for or against post-lumpectomy radiation for patients with DCIS. Rather, it provides a guide regarding expected local recurrence, of either in situ or invasive disease, when radiation therapy is omitted. Established studies demonstrating the reduced risk of in-breast recurrence due to the addition of radiation [30] should be included in the treatment discussion with patients. It also does not stratify recurrence risk into risk of only in situ or of invasive disease, the latter of which is associated with potential mortality and need for more aggressive treatment, such as lymph node evaluation, extended radiation fields, chemotherapy, and biologic therapy. Such information would be particularly informative for guiding treatment decisions. In addition, the DCIS Score does not predict response to post-lumpectomy radiation. Indeed, a trial evaluating the utilization of the DCIS Score on outcomes would be of great interest.

Although the DCIS Score is associated with in-breast recurrence, the test cannot explain the behavior of individual DCIS cases: why does some DCIS remain indolent, and why do other cases progress to invasive disease? With this information, overtreatment may truly be prevented.