A team of investigators in Japan used a personalised assay for circulating tumour DNA (ctDNA) levels in plasma to monitor minimal residual disease following surgery in patients with resectable colorectal cancer. Their subgroup analysis found statistically significant associations between ctDNA levels and the pathological disease stage. No association with RAS, BRAF V600E and MSI was found. The findings from a prospective observational study GALAXY in CIRCULATE-Japan were presented at the ESMO World Congress on Gastrointestinal Cancer 2021, held from 30 June to 3 July 2021.
While explaining the rationale leading to this investigation, Dr. Hiromichi Shirasu of the Division of Gastrointestinal Oncology, Shizuoka Cancer Centre in Sunto-Gun, Japan noted that identifying minimal residual disease after curative surgery by analysis of ctDNA could facilitate the personalisation of adjuvant therapies and early intervention in patients with colorectal cancer.
Dr. Shirasu and colleagues conducted the CIRCULATE-Japan adaptive platform, which comprised a large-scale patient-screening registry, GALAXY, to monitor molecular residual disease status. GALAXY was followed by two ctDNA-guided phase III trials, VEGA and ALTAIR, which aimed to refine adjuvant therapy using ctDNA analysis of minimal residual disease in patients with resectable colorectal cancer.
These studies evaluated the plasma ctDNA status prior and after surgery periodically for up to 2 years. The investigators used SignateraTM, a personalised, tumour-informed ctDNA assay that was designed to track 16 patient-specific somatic variants based on whole-exome sequencing of the tumour tissue and matched normal blood. Median ctDNA levels were measured and expressed as mean tumour molecules per milliliter of plasma. RAS and BRAFV600E mutations, as well as the mismatch repair status in tumour tissue were assessed using polymerase chain reaction-based methods. The investigators also evaluated the association of ctDNA status with tumour biomarkers.
Preoperative ctDNA levels were found to be significantly associated with pathological disease stage
As of 28 February 2021, the GALAXY study enrolled 1,236 patients with 808 patients being included in this analysis. Preoperative baseline ctDNA was detected in 799 patients. ctDNA was available in 797 patients 4 weeks after operation, in 531 patients 12 weeks after operation, and in 263 patients 24 weeks after operation. Stage I-III colorectal cancer had 654 patients and 154 patients had stage IV cancer.
In this cohort, 347 (43%) patients had RAS mutations and 55 (7%) had BRAF V600E mutations. In addition, 71 (9%) patients had microsatellite instable-high (MSI-H) status.
Among 65, 280, and 301 patients with pathological stages I, II, and III, preoperative ctDNA was detected in 50 (77%), 267 (95%), and 288 (96%) patients, respectively.
The investigators found that pT was significant for preopoperative ctDNA positivity, while pN was significant for postoperative positivity at 4 weeks after surgery.
No association with RAS, BRAF V600E and MSI was found.
Longitudinal ctDNA positivity at postoperative week 4, 12 and 24 was significantly associated with inferior disease-free survival (DFS) with hazard ratio (HR) of 46.8. Sensitivity of relapse detection was 93.1%.
Positivity at postoperative week 4 was significantly associated with inferior DFS with HR 19.5 overall, and HR 24.4 in pathologic stage I-III, indicating it is a suitable time point for ctDNA-based adjuvant study.
Multivariate analysis showed that ctDNA was the only significant factor beyond known prognostic factors, HR 17.1.
Six-months DFS rate in ctDNA negative patients in overall and pathologic stage I-III were ≥99%, showing unprecedented good prognosis.
Dr. Shirasu concluded that further investigation whether ctDNA status could become new surrogate endpoint beyond DFS is warranted.
Funding from the Japan Agency for Medical Research and Development was disclosed for this study.
Dr. Jonathan Loree of the BC Cancer, University of British Columbia in Vancouver, Canada who discussed the study data said that minimal residual disease assays provide valuable prognostic information, but it is not yet clear if they are predictive. He questioned on how do we best intercept a positive test and are we comfortable de-escalating treatment?
Funding from the Japan Agency for Medical Research and Development was disclosed.
O11 - Shirasu H, Taniguchi H, Watanabe J, et al. Monitoring molecular residual disease by circulating tumor DNA in resectable colorectal cancer: Molecular subgroup analyses of a prospective observational study GALAXY in CIRCULATE-Japan. ESMO World Congress on Gastrointestinal Cancer 2021 (30 June - 3 July 2021).