Mutational analyses, including tumor tissue somatic mutation or circulating tumor/cell-free DNA analyses, are increasingly used for prognostication and treatment selection for patients with colorectal cancer (CRC).
Determination of KRAS, BRAF, and MSI status is standard of care. Extended molecular profiling has proven advantageous for prognostication in patients with metastatic CRC, particularly patients with resectable CLM.
For example, mutations in KRAS, BRAF, TP53, SMAD4, and FBXW7 are associated with worse survival outcomes following hepatectomy for CLM.1-5
The pathway-centric risk score is valuable when making decisions with patients regarding hepatectomy for CLM, which takes multiple deleterious mutations into account and stratifies patients into three prognostic groups.6 However, more dynamic and precise biomarkers are urgently needed to guide treatment for all stages of CRC.
Circulating tumor DNA (ctDNA) has rapidly emerged as a promising dynamic biomarker to aid with clinical decision-making for CRC. Detection of ctDNA following completion of intended treatment, termed minimal residual disease (MRD), is associated with an extremely high risk for disease recurrence.7 Currently, vague clinical and pathologic factors guide many treatment decisions for CRC. However, would a seismic shift in treatment patterns occur if we could detect occult disease hiding in the shadows?
Localized Colorectal Cancer
Two recent studies have supported use of ctDNA for guiding adjuvant chemotherapy treatment (ACT) decisions after resection of stages II and III—DYNAMIC and GALAXY in CIRCULATE-Japan.
In DYNAMIC, Tie and colleagues randomized patients to standard (ACT for high-risk features) or ctDNA-guided management (ACT if ctDNA detected postoperatively).8 Recurrence-free survival (RFS) at 2 years was equivalent (93.5% and 92.4%, respectively), while ACT use was nearly halved (15% versus 28%), indicating ctDNA evaluation may reduce unnecessary chemotherapy.
A research group from Japan initiated a new type of adaptive platform trials to evaluate the clinical benefits of ctDNA analysis and refine precision adjuvant therapy for resectable colorectal cancer. CIRCULATE-Japan comprises three clinical trials, including the GALAXY study. This trial reported an observational cohort of stage I–III colon cancer patients stratified by postoperative ctDNA status at 4 and 12 weeks.9 Six-month disease-free survival was 98%–100% in patients with no detectable ctDNA at either timepoint (n = 618) or converted to negative with ACT (n = 58) but was only 45% if ctDNA was detectable at both timepoints (n = 78) and 59% if ctDNA became positive 12 weeks postoperatively (n = 32). In addition, ctDNA clearance was higher in patients who received ACT (84% versus 34%).
Four ongoing randomized trials are investigating ACT based on postoperative ctDNA. In the Phase II/III Study of COBRA trial, low-risk stage II patients are randomized to observation versus ctDNA-guided ACT.10
In the CIRCULATE-US trial, high-risk stage II and all stage III patients are randomized to standard-of-care ACT versus observation if they are ctDNA-negative or standard-of-care ACT versus addition of irinotecan if they are ctDNA-positive.11
The VEGA and ALTAIR trials in Japan are randomized observation versus oxaliplatin-based ACT if patients are ctDNA-negative or observation versus trifluridine/tipiracil for patients who are ctDNA-positive but imaging-negative after standard-of-care ACT.12 Together, these trials will provide detailed insight into incorporating ctDNA in clinical decision-making after surgery, thus minimizing side effects from chemotherapy while maximizing cancer-free survival.
Resectable Colorectal Liver Metastases
Like MRD for patients with localized CRC, ctDNA-detection following curative-intent hepatectomy for CLM is associated with oncologic outcomes. In a study of 48 patients undergoing curative-intent hepatectomy with paired pre- and postoperative ctDNA analyses, 38% of patients were ctDNA-positive postoperatively, and these patients had an RFS of only 7.5 months compared to 33 months for those who were ctDNA-negative.13
Subsequently, a tumor comutation in RAS+TP53 was associated with ctDNA detection within 180 days of curative-intent hepatectomy, highlighting patients at high risk for MRD.14 Moreover, serial sampling of ctDNA postoperatively may refine prognostication.15
Patients without detectable ctDNA after curative-intent hepatectomy for CLM are at much lower risk for at least early recurrence, however, these patients do recur. Reinert and colleagues reported a 56% recurrence rate for patients ctDNA-negative postoperatively across their study and, interestingly, usually in the lungs. Lung recurrences were more than 15 times more likely in patients who were ctDNA-negative postoperatively.16 This has significant implications for surveillance following hepatectomy for CLM.
Ongoing clinical trials evaluate the impact of ctDNA-guided treatment strategies for patients with CLM with the objective of better identifying patients who may benefit from chemotherapy, as well as those who may not.
For example, the REACT-CLM trial (NCT05062317) includes patients who undergo curative-intent hepatectomy for CLM after at least four cycles of preoperative chemotherapy. Patients undergo ctDNA analysis for MRD 2–6 weeks postoperatively and, if positive, continue intensive chemotherapy. However, those without MRD de-escalate postoperative chemotherapy or enter surveillance at the discretion of their care team.
With a primary endpoint of 12-month recurrence-free survival, the hypothesis is that those without detectable ctDNA can safely de-escalate postoperative chemotherapy, which has profound implications for quality of life and survivorship.
