SAN ANTONIO – The presence of circulating tumor DNA (ctDNA) and to a lesser degree circulating tumor cells (CTCs) following neoadjuvant chemotherapy for triple-negative breast cancer (TNBC) augurs poor prognosis, but it’s still unclear how this information can inform therapy.
Neil Osterweil/MDedge News
Dr. Milan Radovich
An analysis of plasma samples from 142 women enrolled in a trial that compared outcomes of genomically directed therapy with those of physician’s choice of a regimen for treatment of TNBC showed that, at a median follow-up of 17.2 months, detection of ctDNA was significantly associated with worse distant disease-free survival and the risk of relapse was especially high for patients with both detectable ctDNA and CTCs, reported Milan Radovich, PhD, from Indiana University in Indianapolis.
“Patients with TNBC at high clinical risk of relapse due to incomplete response to neoadjuvant chemotherapy can be further risk-stratified based on the presence of [minimal residual disease] as determined by ctDNA,” he said at the San Antonio Breast Cancer Symposium.
As previously reported, the presence of detectable ctDNA after surgery for breast cancer is a significant predictor of inferior disease-free survival (DFS), but the findings, while clinically valid, are not yet applicable to clinical care, Dr. Radovich acknowledged in a briefing shortly before his presentation of the data in a general session.
As part of the BRE12-158 trial, Dr. Radovich and colleagues conducted a prospective analysis of patients with early-stage TNBC who had significant residual disease at the time of surgery, despite having undergone neoadjuvant chemotherapy. The tumors were genomically sequenced and the patients were then randomized to receive either genomocially directed therapy or the treating physician’s choice of a regimen.
Of 177 evaluable patients, a total of 142 had successful ctDNA results and 123 had successful CTC testing. The ctDNA samples were sequenced using a liquid assay for 70 oncogenes commonly mutated in cancer. CTCs were isolated from peripheral blood with a positive-selection microfluidic device. CTC-positivity was defined at detection of one or more cells.
The investigators found that ctDNA was significantly associated with distant DFS, with a median of 32.5 months for ctDNA-positive patients versus median not reached for ctDNA-negative patients. The respective 2-year distant DFS rates were 56% and 81%, respectively. The hazard ratio for worse distant DFS among ctDNA-positive patients was 2.99 (P = .0055).
Similarly, DFS was significantly inferior among ctDNA-positive patients, with a median of 22.8 months versus not reached for ctDNA-negative patients, and median 2-year DFS rates of 50% and 76%. The HR for worse survival in ctDNA-positive patients was 2.67 (P = .0069).
Overall survival (OS) was also worse for ctDNA-positive patients; although the median OS was not reached in either positive or negative patients, the 2-year OS rates were 57% among ctDNA-positive patients and 80% among negative patients. The HR for death among the ctDNA-positive patients was 4.16 (P = .0024).
CTC detection, in contrast, showed a trend toward a significant association with clinical outcomes, but neither distant DFS, DFS, nor OS measures achieved statistical significance.
However, when ctDNA and CTC were combined, the results showed significantly inferior distant DFS for patients positive for both markers, compared with that of patients who were ctDNA positive but CTC negative (median, 32.5 months vs. not reached; 2 year distant DFS, 52% vs. 89%; HR, 5.29; P = .0095).
Similarly, this group of patients had worse DFS (median, 20.8 months vs. not reached; 2-year DFS, 43% vs. 80%; HR, 3.15; P = .0370) and worse OS (median, OS not reached in either group; 2-year OS, 51% vs. 76%; HR, 8.60; P = .0074).
To take advantage of these data, the investigators are preparing to launch a phase 2 trial (BRE18-334) in which patients with residual disease are randomized by ctDNA status (positive or negative) to either genetically directed therapy or therapy with no genomic target. The patients randomized to genomically directed therapy will be treated according to mutation category.
For example, capecitabine will given plus either a PARP (poly [ADP-ribose] polymerase) inhibitor for patients with mutations in DNA repair pathways, atezolizumab (Tecentriq) for patients with mutations in the immunotherapy pathway, ipatasertib for mutations in the P13K/AKT/mTOR pathway, or a PARP inhibitor plus atezolizumab for patients with mutations in the both DNA repair immunotherapy pathways.
At the briefing, moderator Virginia Kaklamani, MD, DSc, from University of Texas Health San Antonio, commented that the results of the study can inform clinical research, but “we don’t have predictive data as to how we’re going to act on this and improve patient outcomes. What we need is exactly the trial that the investigators have designed to see whether these results can then further help us in treating our patients and improving their outcomes.”
Neil Osterweil/MDedge News
Dr. Brian Schneider
“Should we be doing this on Monday in clinic, and the answer is no,” Dr. Radovich agreed. “What we don’t know is if having this information is really going to improve outcomes.”
In an interview, senior author Bryan P. Schneider, MD from Indiana University School of Medicine put it this way: “I’ve always believed that you don’t order a test until you know what to do with it,” he said.
The study was funded by the Vera Bradley Foundation for Breast Cancer and the Indiana University Grand Challenge Precision Health Initiative. Dr. Radovich and Dr. Schneider reported no relevant disclosures. Dr. Kaklamani disclosed speakers bureau activities and consulting for multiple companies, as well as research funding from Eiasi.
SOURCE: Radovich M. SABCS 2019, Abstract GS5-02.