Validation of the 12-gene colon cancer recurrence score in NSABP C-07 as a predictor of recurrence in patients with stage II and III colon cancer treated with fluorouracil and leucovorin (FU/LV) and FU/LV plus oxaliplatin

Abstract

Purpose: Accurate assessments of recurrence risk and absolute treatment benefit are needed to inform colon cancer adjuvant therapy. The 12-gene Recurrence Score assay has been validated in patients with stage II colon cancer from the Cancer and Leukemia Group B 9581 and Quick and Simple and Reliable (QUASAR) trials. We conducted an independent, prospectively designed clinical validation study of Recurrence Score, with prespecified end points and analysis plan, in archival specimens from patients with stage II and III colon cancer randomly assigned to fluorouracil (FU) or FU plus oxaliplatin in National Surgical Adjuvant Breast and Bowel Project C-07.

Methods: Recurrence Score was assessed in 892 fixed, paraffin-embedded tumor specimens (randomly selected 50% of patients with tissue). Data were analyzed by Cox regression adjusting for stage and treatment.

Results: Continuous Recurrence Score predicted recurrence (hazard ratio for a 25-unit increase in score, 1.96; 95% CI, 1.50 to 2.55; P < .001), as well as disease-free and overall survival (both P < .001). Recurrence Score predicted recurrence risk (P = .001) after adjustment for stage, mismatch repair, nodes examined, grade, and treatment. Recurrence Score did not have significant interaction with stage (P = .90) or age (P = .76). Relative benefit of oxaliplatin was similar across the range of Recurrence Score (interaction P = .48); accordingly, absolute benefit of oxaliplatin increased with higher scores, most notably in patients with stage II and IIIA/B disease.

Conclusion: The 12-gene Recurrence Score predicts recurrence risk in stage II and stage III colon cancer and provides additional information beyond conventional clinical and pathologic factors. Incorporating Recurrence Score into the clinical context may better inform adjuvant therapy decisions in stage III as well as stage II colon cancer.

Trial registration: ClinicalTrials.gov NCT00004931.

Conflict of interest statement

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

Figures

Fig 1.

Flowchart showing patient and sample accounting in National Surgical Adjuvant Breast and Bowel Project (NSABP) C-07; 4.5% of patients eligible for the study did not have death certificates.

Fig 2.

Relationship between the continuous Recurrence Score (RS) and 5-year recurrence risk by stage and treatment in National Surgical Adjuvant Breast and Bowel Project C-07. Thick lines represent fluorouracil (FU) –treated patients, thin lines represent FU + oxaliplatin–treated patients. Blue, gold, and gray colors represent stage II, IIIA/B, and IIIC, respectively. A rug plot depicting the distribution of RS is included at the bottom of the figure, and an estimated normal distribution of scores is provided below. The proportional hazards assumption held (P = .20 for the test of nonzero slope of Schoenfeld residuals v time). The relationship between continuous RS and the log hazard of recurrence was linear (P = .84 for the test of nonlinearity).

Fig 3.

Recurrence-free interval for (A) all patients and (B) stage II, (C) stage IIIA/B, and (D) stage IIIC patients by Recurrence Score (RS) groups and treatment in National Surgical Adjuvant Breast and Bowel Project C-07. Thick lines represent fluorouracil (FU) –treated patients, and thin lines represent FU + oxaliplatin (FU + Ox) –treated patients. Gray, gold, and blue colors represent low, intermediate (int), and high RS groups, respectively. Across all patients (A), FU-treated patients in the high RS group had higher risk of recurrence compared with patients treated with FU + Ox (log-rank P = .016), whereas patients in the low RS group had similar recurrence risks (log-rank P = .98). Numbers of deaths without recurrence were 20, 30, and 12 in stage II, IIIA/B, and IIIC, respectively.

Fig 4.

Distribution of the Recurrence Score (RS) by levels of covariates in National Surgical Adjuvant Breast and Bowel Project C-07: (A) stage, (B) nodes examined, (C) age, (D) mismatch repair (MMR) status, (E) T stage, and (F) tumor grade. P values are from an analysis of variance test for association of the continuous RS with the levels of the covariate. Gray horizontal lines represent cut points for low and high RS groups (RS = 30 and 41, respectively); percentage of patients in each RS group at each level of the covariate is shown.

Fig A1.

Recurrence-free interval by stage and treatment in National Surgical Adjuvant Breast and Bowel Project C-07. Thick lines represent fluorouracil (FU) –treated patients, and thin lines represent FU + oxaliplatin (FU + Ox) –treated patients. Blue and gold colors represent stage II and III, respectively.

Fig A2.

Cumulative incidence plots for recurrence-free interval for (A) stage II, (B) stage IIIA/B, and (C) stage IIIC patients by Recurrence Score (RS) groups and treatment in National Surgical Adjuvant Breast and Bowel Project C-07. Thick lines represent fluorouracil (FU) –treated patients, and thin lines represent FU + oxaliplatin (FU + Ox) –treated patients. Blue, gold, and gray colors represent low, intermediate (int), and high RS groups, respectively.