Aims
Detection of ctDNA can identify patients with minimal residual disease (MRD) in curatively-treated CRC. Highly sensitive NGS-based assays can have comparable sensitivity to digital PCR, with broader coverage and no requirement for prior tumour sequencing. The ASpirin for Dukes C and high-risk Dukes B COLorecTal cancer trial (ASCOLT, NCT00565708) is a randomised, double-blinded, phase III fully accrued study investigating adjuvant aspirin benefit. In the Australian and New Zealand cohort, serial plasma samples were prospectively collected at study enrolment, 6 and 12 months (m), along with baseline tumour tissue. In this pilot, we evaluated the utility of an ultra-sensitive ctDNA NGS assay (SafeSEQ, Sysmex Inostics).
Methods
Extracted cell-free DNA was analysed using a SafeSEQ CRC MRD 14-gene assay: AKT1, APC, BRAF, CTNNB1, ERBB3, FBXW7, KRAS, NRAS, PIK3CA, POLE, PPP2R1A, RNF43, SMAD4 and TP53. Tumour sequencing used an amplicon-based 59-gene custom panel. ctDNA was correlated with tumour sequencing and recurrence data.
Results
Twenty-nine plasma samples (median 3.6mL) from the first 10 patients were analysed (median age 65; 2 stage II, 8 III). All received adjuvant 5-Fluorouracil-based chemotherapy. Median follow-up was 59.9m, with 3 known recurrences. Two of these patients had detectable ctDNA: 1 at 6m (KRAS G12V, lung metastases) and the other at 12m (KRAS G13D/ SMAD4 L540P, lung and nodal metastases). Cancer recurrence was detected by CT 16.5m after and within 1m, respectively, of ctDNA positivity. Both plasma mutation profiles were concordant with tumour analysis. A third patient recurred 59.2m after commencement (peritoneal metastases) without detectable ctDNA. CEA did not predict recurrence in any of the 3 patients.
Conclusions
In this pilot cohort, 2 of 3 patients who recurred had detectable ctDNA using a SafeSEQ assay (concordant with tumour sequencing results). A third patient with late recurrence did not have detectable ctDNA. Analyses in the larger cohort (n=368) are underway.