Predictors of DCIS Recurrence and Risk of Invasive Cancer: Overview of the Field and Current Challenges

Lawrence J. Solin, MD, FACR; University of Pennsylvania

View Presentation (PDF)

Currently, the predictors for local recurrence after breast conservation treatment for DCIS are based on traditional patient, tumor, and treatment characteristics, and are not fully adequate for patients and physicians to make optimal treatment decisions. Therefore, new biologically-based predictors for tailored treatments are needed. Risk factors for local recurrence come from many areas, including patient characteristics (e.g., age), tumor characteristics (e.g., size, margins, pathology), biology (e.g., hormone receptor status), and treatment (e.g., surgery, radiation, tamoxifen, hormones). Tailoring treatment to these risk factors may be important in successfully treating DCIS. Because DCIS has about a 97-98 percent survival rate, the endpoint to be considered for successful treatment is local recurrence, both invasive local recurrence and total local recurrence. Unfortunately, the parameters to predict local recurrence are crude, and the selection of patients for treatment is based on these crude parameters.

Researchers are continuing to collect and analyze tissue blocks to examine molecular predictors of local recurrence. A number of biologic factors have been evaluated to attempt to correlate with local recurrence, including ER (estrogen receptor), PR (progesterone receptor), p53, HER-2/neu, Ki-67, p21, and BCL-2. However, none has consistently been correlated with local control. Combinations of specific factors have the potential to be more specific than single factors in predicting local recurrence.

Models are in development to estimate the outcomes for DCIS patients who have undergone breast conserving surgery. These data can inform and educate patients for decision-making to determine treatment. Such programs may be helpful for patients in dealing with the current challenges of treating DCIS, which include developing predictive factors to tailor treatment for individual patients and adding biologically based factors to current models based on patient and tumor factors for tailored treatment.