SAN ANTONIO (FRONTLINE MEDICAL NEWS) – A biological risk signature can help guide decisions about use of adjuvant radiation therapy in patients with ductal carcinoma in situ (DCIS), suggests a validation study reported at the San Antonio Breast Cancer Symposium.

Radiation therapy reduces the 10-year risk of any ipsilateral recurrence in this population by about 50% as established in a large overview of trials ( J Natl Cancer Inst Monogr. 2010;2010:162-77 ), noted lead investigator Fredrik Wärnberg, MD, PhD, of Uppsala Academic Hospital, Uppsala University, Sweden. But factors such as tumor size, grade, and margins have not been helpful in identifying patients most likely to benefit.

He and his colleagues validated a biological risk signature (DCISionRT; PreludeDx) among 584 patients with pure primary DCIS treated on the SweDCIS trial. The trial randomized patients who had undergone breast-conserving surgery to receive radiation therapy or not, and then followed them for 20 years.

The risk signature incorporates four clinicopathologic factors and seven immunohistochemically assessed biomarkers of hormone receptor status, HER2 status, stress response, and proliferation. Possible scores range from 0 to 10, and are split into categories of low risk (0 to 3) and elevated risk (greater than 3). “To me, the magic of this signature is that it is nonlinear. Each factor can be dependent on the value of other factors in the model,” Dr. Wärnberg said.

Results of the validation study showed that among the 506 patients who had clear margins after surgery, radiation therapy significantly reduced the 10-year risk of invasive recurrence in those with an elevated risk score by more than three-fourths, but not in those with a low risk score.

“The biologic risk signature … correlated to risk. It’s prognostic, that’s nothing new,” he summarized. “More interestingly, it was also predictive for radiotherapy benefit. Not all patient groups had the same benefit from radiation therapy. In the low-risk group, there wasn’t any significant benefit from radiation therapy for invasive recurrences. But in the elevated risk group, the radiation therapy benefit was twice as high as expected – about a 76% relative risk reduction with radiotherapy for invasive recurrences.”

Study details

Main results from the SweDCIS trial, previously reported ( J Clin Oncol. 2014;32:3613-8 ), showed that adjuvant radiation therapy reduced recurrences, yielding a 12% absolute reduction in risk of ipsilateral recurrence (10% for in situ recurrences and 2% for invasive recurrences).

For the validation study, Dr. Wärnberg and his colleagues were able to obtain tissue and biological signature results, blinded to patient outcome, for 56% of the original trial population. About half of patients each were determined to have low risk scores and elevated risk scores.

Among the 506 patients with clear margins, the score, analyzed as a continuous variable, was associated with risk of any (in situ or invasive) ipsilateral recurrence during follow-up (hazard ratio, 1.49 per 5-unit increase; P = .038).

In a multivariate model, receipt of radiation therapy was associated with a 52% relative reduction in 10-year risk of any ipsilateral recurrence for those with a low-risk score (HR, 0.48; P = .04) and a greater 69% relative reduction for those with an elevated risk score (HR, 0.31; P less than .001).

Radiation therapy did not significantly reduce the risk of ipsilateral invasive recurrence in the low risk group (HR, 0.84; P = .70), but it did in the elevated risk group (HR, 0.24; P = .012).

These findings essentially mirrored those of a 2015 validation study in a separate cohort of 526 patients from Uppsala University Hospital and the University of Massachusetts, according to Dr. Wärnberg. “We found highly consistent data with these two different sets,” he said.

Analyses additionally showed that radiation therapy reduced the 10-year risk of an invasive breast cancer recurrence by an absolute 1% for patients with low risk scores (not significant) but by an absolute 9% for patients with elevated risk scores (P = .012).

Dr. Wärnberg disclosed that he had no relevant conflicts of interest. The study was funded in part by PreludeDx.

SOURCE: Warnberg et al., SABCS Abstract GS5-08