Assessment of molecular minimal residual disease (MRD) could potentially be leveraged, for example, to avoid overtreatment of early stage breast cancer through individualization of local and/or adjuvant therapy based on response to neoadjuvant therapy (NAT). In this context, however, >90% of women have no MRD detected using current liquid biopsy assays, regardless of response to NAT, thus precluding reliable risk stratification. The newly developed targeted digital sequencing (TARDIS) platform for personalized detection and quantitation of circulating tumour DNA (ctDNA) might have the required sensitivity.

Combining the strengths of PCR and ligation-based sequencing assays, TARDIS enables screening of plasma cell-free DNA for several patient-specific founder mutations stringently identified via whole-exome sequencing of tumour DNA. After extensive technical validation of TARDIS using reference samples — revealing a sensitivity of at least 53.1% at a variant allele fraction (VAF) of 0.003% in a clinically feasible volume of plasma (4 ml), with almost 100% specificity — the assay was applied to 80 plasma samples from 33 women with stage I–III breast cancer, with 6 to 115 mutations assessed per patient.

ctDNA was detected in pretreatment samples from all but one patient, at a median VAF of 0.11%, and in post-NAT samples from 17 of 22 women, at a median VAF of 0.016% — a median 84% decrease. Notably, 5 of 9 women with a pathological complete response (pCR) had MRD detected, compared to 12 of 13 without a pCR. Moreover, the ctDNA AFs were almost sixfold lower in those with a pCR (median 0.003% versus 0.018%); the median decrease in ctDNA concentrations was 96% versus 77%. Accordingly, ctDNA fraction enabled prediction of pathological residual disease with an area under the curve of 0.83.

Importantly, development of the patient-specific assays typically took 3–4 weeks from the time of diagnostic biopsy sampling, which is compatible with the timeframe of post-NAT decision-making. The authors acknowledge, however, that prospective studies with larger cohorts and long-term follow up are needed to establish the optimal diagnostic thresholds and the clinical validity and utility of TARDIS.