A new gene classifier can predict the risk of cancer cells recurring or progressing

Mapping a molecular atlas of ductal carcinoma in situ (DCIS) has made a team of researchers a major advance toward distinguishing whether early breast cancers develop into invasive cancers or remain stable.

By analyzing samples from patients who had surgery to remove areas of ductal carcinoma in situ, the team identified 812 genes associated with cancer development. Using this gene classifier, they were then able to predict the risk of cancer cells recurring or progressing.

The study, published this week in the journal Cancer Cell, was led by E. Shelley Hwang, MD, of the Duke Cancer Institute, and Rob West, MD, PhD, of Stanford University Medical Center. Their work is part of the Human Tumor Atlas Network under the Moonshot Initiative funded by the National Cancer Institute.

“There has been a long-standing debate about whether ductal carcinoma in situ is a cancer or a high-risk condition,” Hwang said. “In the absence of a way to make that decision, we are currently treating everyone with surgery, radiation, or both.

“Ductal carcinoma in situ is diagnosed in more than 50,000 women annually, and about a third of those women undergo mastectomy, so we are increasingly concerned that we may be overtreating many women,” Huang said. “We need to better understand the biology of ductal carcinoma in situ, and that’s what our research was designed for.”

Hwang, West and colleagues analyzed 774 ductal carcinoma in situ samples from 542 patients at an average of 7.4 years after treatment. They identified 812 genes associated with recurrence within five years of treatment.

The genetic classifier was able to predict both cancer recurrence and invasive progression, as progression appears to depend on a process that requires interactions between invasive DCIS cells and the unique features of the tumor environment.

Hwang said most DCIS cancers analyzed in the study were identified as being at low risk of cancer progression or recurrence — a factor that underscores the need for an accurate predictive model that can be used during clinical visits to guide care.

We’ve made great progress in our understanding of DCIS, and this work gives us a real path forward toward being able to personalize care by expanding the range of treatments for the risks of cancer progression. The real goal is to reduce treatment-related harms without compromising results, and we’re excited to get closer to achieving that for our patients with ductal carcinoma in situ. “

E. Shelley Hwang, MD, Duke Cancer Institute

In addition to Hwang and West, study authors include co-principal investigator Carlo Malli, PhD, of the Arizona State School of Life Sciences, and Graham Colditz, PhD, of Washington University in St. Louis, Breast Cancer Atlas Center, as well as collaborators from 12 other institutions. As part of the Transitional Breast Cancer Consortium.

The study is part of the National Cancer Institute’s Human Tumor Atlas Network Consortium, which is part of the National Institutes of Health (R01 CA185138-01, U2C CA-17-035, UO1 CA214183, R01CA193694). Other funding support was from the Department of Defense (BC132057); Breast Cancer Research Foundation (19-074, 19-028, 18-006); PRECISION CRUK Grand Challenge (AEI RYC2019-026576-I); La Caixa Foundation (LCF/PR/PR17/51120011); Lundbeck Foundation (R288-2018-35); Danish Cancer Society (R229-A13616); and Susan J. Komen.