Circulating Tumor DNA as a Marker of Recurrence Risk in Stage III Colorectal Cancer: The α-CORRECT Study

Abstract

Background and objectives: Identification of colorectal cancer (CRC) patients at high risk of recurrence could be of substantial clinical use. We evaluated the association of ctDNA status, using a tumor-informed assay, with recurrence-free survival (RFS).

Methods: Stage III CRC patients were enrolled between 2016 and 2020. Tumor tissue and serial (every 3 months for years 1-3, biannually for years 4-5) blood samples were collected. Utilizing whole-exome sequencing and selection of 50-200 variants for tumor informed assays, ctDNA status was determined using plasma cell-free DNA.

Results: Of 137 patients enrolled, 124 with 1029 ctDNA results were included in the analyses. Median follow-up was 4.8 years. Plasma ctDNA status was strongly associated with risk of recurrence during the surveillance period (hazard ratio (HR) 49.6, 95% CI: 16.6-148.3; p < 0.0001), and at the postsurgical (HR 9.6, 95% CI: 3.2-29.5) and postdefinitive therapy timepoints (HR: 16.7, 95% CI: 6.9-40.3). The estimated 3-year RFS for ctDNA positive and ctDNA negative patients were, respectively, 54.5% and 96.1% after surgery, and 18.2% and 90.0% after definitive therapy. Multivariable analysis indicated ctDNA but not CEA was strongly prognostic for recurrence.

Conclusions: Our tumor-informed ctDNA assay was strongly prognostic for recurrence in patients with stage III colorectal cancer at all timepoints.

Keywords: colorectal cancer; ctDNA; molecular residual disease (MRD); tumor‐informed.

Figure 1

Study overview showing sampling timepoints (A) and number of patients with each outcome (B). (A) Blood draws utilized for the current study are in red. Green boxes represent the three timepoints used for ctDNA analyses. The surveillance period timepoint included ≥ 1 sample, while postsurgical (PS) and post‐definitive therapy (PDT) timepoints included a single sample. Samples collected during adjuvant therapy or after recurrence were not utilized and are shown in gray. (B) CONSORT diagram.

Figure 2

Kaplan‐Meier recurrence‐free survival (RFS) curves by ctDNA status. (A) At least one positive ctDNA result versus all negative ctDNA results during the surveillance time period. (B) Positive versus negative ctDNA at the postsurgical (PS) timepoint. (C) Positive versus negative ctDNA at the post‐definitive therapy (PDT) timepoint.

Figure 3

Kaplan‐Meier recurrence‐free survival (RFS) curves by ctDNA status using the adjusted criteria for ctDNA status determination. (A) At least one positive ctDNA result versus all negative ctDNA results during the surveillance time period. (B) Positive versus negative ctDNA at the postsurgical (PS) timepoint. (C) Positive versus negative ctDNA at the post‐definitive therapy (PDT) timepoint.

Figure 4

Clearance of ctDNA and recurrences following adjuvant chemotherapy. (A) Sankey plot showing the 22 patients who were ctDNA positive at the postsurgical (PS) timepoint and had a post‐definitive therapy (PDT) timepoint ctDNA result. Thirteen of these patients showed clearance, defined as being ctDNA positive at the PS timepoint and ctDNA negative at the PDT timepoint, and nine did not show clearance. The ctDNA results following the PDT timepoint were either consistently positive, consistently negative, mixed–having both positive and negative results during the surveillance period, or absent (no ctDNA result). (B) Kaplan‐Meier curves for recurrence‐free survival (RFS) by ctDNA clearance status: positive at PS and PDT (no clearance) versus positive at PS and negative at PDT (clearance). (C) Kaplan‐Meier curves for RFS for patients who were ctDNA positive at the PS timepoint, had a PDT timepoint sample and at least one additional sample after the PDT timepoint, by ctDNA status during surveillance: persistent ctDNA positive, persistent ctDNA negative, or mixed ctDNA results.