Conference Coverage

Circulating microRNAs may predict breast cancer treatment response

 

Key clinical point: Circulating microRNAs may predict treatment response in HER2-positive breast cancer patients treated with neoadjuvant therapy.

Major finding: A total of 30 microRNAs and 6 microRNA signatures were found to predict pathologic complete response in a “training population,” and the predictive ability of 4 of those was confirmed in the testing set.

Data source: An analysis of data from the randomized phase III NeoALTTO trial.

Disclosures: Dr. Di Cosimo reported having no disclosures.


 

AT SABCS 2016

– Circulating microRNAs may predict treatment response in HER2-positive breast cancer patients treated with neoadjuvant therapy, according to findings from the randomized phase III NeoALTTO trial.

Specifically, four circulating tumor microRNA signatures were found to predict pathologic complete response to specific treatments at specific time points in a “testing set” of patients from the study, Serena Di Cosimo, MD, of Istituto Nazionale dei Tumori, Milan, reported at the San Antonio Breast Cancer Symposium.

Overall results from the trial, which were reported in 2012 (The Lancet. 379[9816]:633-40) showed that the combination of lapatinib and trastuzumab significantly improved rates of pathologic complete response, compared with either drug alone in patients with HER2-positive early breast cancer, and the authors concluded that dual inhibition of HER2 might be a valid approach in such patients in the neoadjuvant treatment setting.

For the current analysis of NeoALTTO data, microRNA from plasma samples from 435 women – 141 treated with lapatinib, 151 treated with trastuzumab, and 143 treated with a combination of the two – were monitored longitudinally, as NeoALTTO “was done thinking about future translational studies, so blood samples were collected prospectively at baseline, after 2 weeks of treatment, at surgery, and at the time of relapse,” she noted.

Potential microRNAs associated with pathologic complete response were identified for each treatment arm and time point, and a multivariate model was used to identify specific signatures and their predictive capability.

A total of 30 microRNAs and 6 microRNA signatures were found to be predict pathologic complete response in a “training population,” and the predictive ability of 4 of those was confirmed in the testing set for lapatinib at baseline (area under the curve [AUC] = .86), lapatinib after 2 weeks (AUC = .71), trastuzumab after 2 weeks (AUC=.81), and lapatinib and trastuzumab after 2 weeks (AUC = .67), Dr. Di Cosimo said.

“Results obtained early post treatment are of special value,” she said, explaining that “women with unfavorable microRNA signatures can be expected to have poor response after just 2 weeks of treatment.”

The findings are of note because a significant proportion of breast cancer patients treated in the neoadjuvant setting do not achieve pathologic complete response and have an increased risk of relapse after surgery. Further, there is a lack of reliable predictors of response to help guide therapy in clinical practice, Dr. Ci Cosimo said.

“This is the first evidence of the potential of circulating microRNAs to discriminate between responsive and unresponsive HER2+ breast cancer patients,” she said.

At present, the signatures, overall, have been associated with pathologic complete response, but not event-free survival. However, one microRNA signature (140-5p) appeared to be associated with event-free survival after 2 weeks among patients treated with trastuzumab, she said, adding that functional studies are ongoing to investigate the biological role of microRNAs, and independent validation studies are planned.

Dr. Di Cosimo reported having no disclosures.

   Comments ()

Recommended for You

News & Commentary

Quizzes from MD-IQ

Research Summaries from ClinicalEdge