Efforts Are Underway to Refine Prognostic Value of ctDNA to Guide Adjuvant Therapy in CRC

Circulating tumor DNA(ctDNA) status at the time of postoperative minimal residual disease (MRD) is a prognostic factor of recurrence for patients with radically resected stage II to IV colorectal cancer (CRC). Standard management calls for adjuvant chemotherapy for all patients following surgical resection; however, investi- gators have sought to refine the role of ctDNA for those who may avoid treatment in the adju- vant setting.^1,2

Recent analyses have shown that the presence of ctDNA alone and in tandem with other mark- ers such as BRAF status, and clinical risk factors may improve the prognostic implications and help guide choice of appropriate candidates for adju- vant treatment.^3-6

Putting data into Context

Retrospective and prospective data leverag- ing ctDNA has guided the initiation of trials to further clarify its prognostic significance. The clearance of ctDNA has also increasingly been used as a surrogate end point leaving investi- gators with a plethora of applications for the liquid biopsy–obtained marker.1 In 2020, the National Cancer Institute Colon and Rectal-Anal Task Force issued a white paper on the applica- tions of ctDNA in CRC treatment research. In its key recommendations the task force wrote, “Current data suggest that detectable ctDNA after surgery and/or completion of adjuvant therapy is strongly associated with a high risk of disease recurrence, suggesting that ctDNA is a robust marker for MRD.”²

The prognostic value of identifying subgroups of patients with CRC for whom standard of care adjuvant chemotherapy may be omitted without detrimental effects on survival has been investi- gated in prospective analyses such as IDEA. In the IDEA analysis, which pooled data from 6 trials to compare adjuvant chemotherapy given 3 months vs 6 months to patients with stage III colon cancer, investigators broke the staging into smaller risk stages to identify the benefit of adjuvant therapy.^5,7 Although the IDEA trial was designed to determine the noninferior- ity of the treatment duration on overall survival, the differences among those classified as having highest-risk disease (T4N2b) and lowest-risk disease (T1N1a) point toward the significance of identifying markers to isolate patients who would benefit most from adjuvant treatment. The 5-year disease-free survival (DFS) rates were 31% vs 89%, respectively.⁵

ctDNA-guided treatment has been evaluated in studies such as the phase 2 DYNAMIC trial (ACTRN12615000381583). Specifically, treatment with adjuvant chemotherapy with oxaliplatin-based or fluoropyrimidine chemother- apy was prompted by ctDNA positivity at 4 or
7 weeks after surgery for patients with stage II colon or rectal adenocarcinoma. Like the design of the IDEA analysis, noninferiority was estab- lished between those who were treated using ctDNA-guided regimens (n = 294) vs standard management (n = 147). The 2 arms were nonin- ferior with patients in the ctDNA-guided arm having a 3-year recurrence-free survival (RFS) rate of 91.7% compared with 92.4% in the standard management arm (HR, 0.96; 95% CI, 0.51-1.82).⁶

Morover, the 3-year RFS rate for those with ctDNA-positive disease (n = 45) was 86.4% compared with 92.5% of patients in the ctDNA-neg- ative cohort (n = 246) who received no treatment. With most patients showing ctDNA negativity, investigators sought to explore efficacy based on clinical risk characteristics in this cohort. In the post hoc analysis, the 3-year RFS rate was 96.7% for patients with low-risk characteris-
tics (n = 156) vs 85.1% for those with high-risk characteristics (n = 89; HR, 3.04; 95% CI, 1.26- 7.34). For patients with T3 tumors (n = 213), the 3-year RFS rate was 94.2% vs 81.3% for those with T4 tumors (n = 33; HR, 2.60; 95% CI, 1.01-6.71). Investigators concluded that adjuvant therapy should not be considered for those who have ctDNA negativity and low-risk disease, which accounted for 11% of patients included in the standard management group.⁶

GALAXY, the observational arm of the ongo- ing CIRCULATE-Japan trial (UMIN000039205), is monitoring the ctDNA status of patients with stage II to IV recurrent CRC who are eligible for complete surgical resection. As part of the larger protocol within GALAXY, following whole- genome sequencing, patients will be treated in 1 of 2 trials based on their ctDNA status. The phase 3 VEGA trial (jRCT1031200006) is comparing capecitabine/oxaliplatin with observation alone in patients with high-risk stage II or low-risk stage III CRC and negative ctDNA status 4 weeks after surgery in GALAXY, and the phase 3 ALTAIR trial (NCT04457297) compares trifluridine/tipira- cil with placebo in patients with resected CRC found to be positive for ctDNA in the GALAXY
tial.^3,4

Findings from the analysis published in Nature Medicine showed that among those who were ctDNA positive at 4 weeks follow- ing surgery, 61.4% (n = 115 of 187) experienced recurrence vs 9.5% (n = 81 of 852) who were ctDNA negative (HR, 10.0; 95% CI, 7.7-14.0; P < .0001). The 18-month DFS rates were 38.4% (95% CI, 31.4%-45.5%) vs 90.5% (95% CI, 88.3%- 92.3%), respectively.³

In updated data from the GALAXY study, presented during the 2023 American Society of Clinical Oncology Breakthrough meeting, the 18-month DFS rate for patients who were postop- eratively ctDNA negative at 4 weeks (n = 1797) was 93.9% (95% CI, 92.5%-95.0%) vs 51.6% (95% CI, 45.2%-57.6%) among patients who were ctDNA posi- tive (n = 286; HR, 12.0; 95% CI, 9.1-15.0; P < .001). Ninety-six patients in who were ctDNA negative experienced a DFS event compared with 130 who were ctDNA positive (TABLE 1).⁴

Additionally, patients who converted to ctDNA-negative status postoperatively (n = 112; HR, 3.5; 95% CI, 1.9-5.8; P < .001), converted to ctDNA-positive status postoperatively (n = 43; HR, 14.5; 95% CI, 8.8-23.8; P < .001), or were persistently ctDNA positive (n = 124; HR, 25.4; 95% CI, 18.3-35.3; P < .001) between 4 to 12 weeks, were all more likely to have disease recurrence compared with those who had ctDNA persistently negative dynamics (n = 1529). The 18-month DFS rates in these subgroups were 82.2% (95% CI, 72.3%-88.9%), 47.4% (95% CI, 30.4%-62.7%), 33.8% (95% CI, 25.4%-42.8%), and 94.9% (95% CI, 93.5%- 96.0%), respectively.⁴

“Our study builds on existing evidence from the recently published, prospective GALAXY study, demonstrating the prognostic value of ctDNA analyzed in more than 2000 patients,” Jun Watanabe, MD, PhD, said during the presen- tation. Watanabe is an associate professor in the Department of Surgery at the Gastroenterological Center at Yokohama City University Medical Center in Japan. He added that the analysis demonstrated that variant data may play a role in determining DFS rates for patients.

The 18-month DFS rates for patients who were postoperatively ctDNA negative at 4 weeks with BRAF wild-type/microsatellite instability–high (MSI-H; n = 88), BRAF V600E–mutant/MSI-H (n = 107), BRAF wild-type/microsatellite stable (MSS; n = 1450), and BRAF V600E–mutant/MSS disease (n = 50) experienced 18-month DFS rates were 100%, 100%, 93.3% (95% CI, 91.7%-94.6%), and 86.3% (95% CI, 71.8%-93.7%), respectively (TABLE 2).4 Using patients with BRAF wild-type/ MSS disease as the reference group, the HRs were 0.09 (95% CI, 0.001-0.61; P = .006), 0.07 (95% CI, 0.001-0.50; P = .002), and 2.06 (95% CI, 0.84-4.23; P = .107) in the BRAF wild-type/MSI-H, BRAF V600E–mutant/MSI-H, and BRAF V600E–mutant/ MSS groups, respectively.⁴

Comparatively, the 18-month DFS rates for those who were postoperatively ctDNA positive at 4 weeks with BRAF wild-type/MSI-H (n = 9), BRAFV600E–mutant/MSI-H (n = 3), BRAF wild-type/ MSS (n = 257), and BRAF V600E–mutant/MSS disease (n = 9) were 66.7% (95% CI, 28.2%-87.8%), 0%, 50.9% (95% CI, 44.1%-57.3%), and 33.3% (95% CI, 7.8%-62.3%). Patients with BRAF wild-type/MSS disease served as the reference group and the HRs were 0.67 (95% CI, 0.21-2.20; P = .488), 7.54 (95% CI, 2.37-24.0; P < .001), and 2.33 (95% CI, 1.03-5.30; P = .043) in the BRAF wild-type/MSI-H, BRAF V600E–mutant/MSI-H, and BRAF V600E–mutant/ MSS groups, respectively.⁴

“Combining BRAF and MSI status with 4-week postoperative ctDNA status increased the predictive value of BRAF and MSI status. Our multivariate analysis for DFS indicated that post- operative ctDNA [positive] status at 4 weeks, BRAF mutational status, and MSI status were signifi- cantly associated with DFS,” Watanabe said.