Detection of minimal residual disease and prediction of recurrence in breast cancer using a plasma-only circulating tumor DNA assay

Abstract

Background: Detection of minimal residual disease (MRD) in early breast cancer (EBC) after curative-intent treatment may identify patients at risk for recurrence. Most circulating tumor DNA (ctDNA)-based MRD assays require knowledge of genomic alterations from tumor tissue. However, tissue availability may be limited in some patients. Here, we evaluated sensitivity and specificity for recurrence detection, using a plasma-only ctDNA MRD assay.

Materials and methods: For this pilot study, 47 plasma samples from 38 EBC patients were collected at 12 or 36 months post-diagnosis or at clinical recurrence. ctDNA presence was determined by a custom bioinformatics classifier that identifies tumor-derived somatic variants and methylation profiles specific to individual cancer types using a 5-Mb next-generation sequencing panel.

Results: ctDNA was detected at or before distant recurrence in 11/14 (79%) patients [sensitivity was 85% (11/13) among samples collected within 2 years from recurrence]. Lead time was evaluable in 4/6 (67%) samples collected before distant recurrence with detectable ctDNA and ranged from 3.4 to 18.5 months. ctDNA was not detected in samples from patients without recurrence (n = 13).

Conclusions: This study demonstrates the feasibility of MRD detection in EBC using a plasma-only multiomic ctDNA-based approach. Larger studies are ongoing to further validate the clinical performance of the assay and demonstrate its applications.

Keywords: MRD; breast cancer recurrence; circulating tumor DNA; ctDNA; early breast cancer; minimal residual disease.

Figure 1

Breast cancer clinical feasibility pilot composition. Forty-seven plasma samples from 38 patients with early-stage breast cancer were collected through the BRandO BiO registry at 12 or 36 months post-diagnosis and/or at the time of clinical recurrence. ^aSelected sequential patients to enrich for recurrences for the purpose of feasibility.

Figure 2

Swimmer plot outlining timing of treatment, relapse, and ctDNA collection. ctDNA, circulating tumor DNA; HER2, human epidermal growth factor receptor 2.

Figure 3

VAF by recurrence status. The lowest estimated ctDNA level was detected in a patient with local recurrence (0.0002%). The median estimated ctDNA level was 0.034% (range 0.026%-5.0%) in 5 samples collected before distant recurrence and 6.5% (range 0.39%-44.3%) in 12 samples collected at the time of distant recurrence. ctDNA, circulating tumor DNA; ND, not detected; VAF, variant allele fraction.