www.nature.com/bjc ARTICLE Clinical Study Association of post-operative CEA with survival and oxaliplatin benefit in patients with stage II colon cancer: a post hoc analysis of the MOSAIC trial Edouard Auclin1,2,3, Thierry André4,5, Julien Taieb1,6, Maria Banzi7, Jean-Luc Van Laethem8, Josep Tabernero9, Tamas Hickish10, Aimery de Gramont5,11 and Dewi Vernerey2,3,5 BACKGROUND: Adjuvant treatment for stage II colon cancer (CC) can be proposed to patients with high-risk disease. Recently, 2.35 ng/mL carcinoembryonic antigen (CEA) was identified as the best cut-off value. This post hoc analysis of the MOSAIC trial assessed post-operative CEA prognostic value for survival outcomes and predictive value for the addition of oxaliplatin to adjuvant treatment. METHODS: Prognostic and predictive values of post-operative CEA in patients with stage II CC were evaluated with Kaplan–Meier survival curves and Cox model with interaction terms. Disease-free survival (DFS) and overall survival (OS) were estimated. RESULTS: Among 899 stage II CC patients, post-operative CEA was available in 867 (96.4%); and 434 (48.65%) had a high-risk stage II disease. The 3-year DFS rate was 88.5% and 78.7% in the ≤ 2.35 ng/mL and > 2.35 ng/mL group, respectively (P = 0.006). Use of oxaliplatin showed survival benefit only in patients with high-risk stage II CC and post-operative CEA > 2.35 ng/ml (interaction term P = 0.09 and 0.03 for DFS and OS). CONCLUSION: CEA is a strong prognostic factor for DFS and OS in stage II CC. In the MOSAIC trial, only high-risk stage II CC patients with post-operative CEA > 2.35 ng/mL benefited from the addition of oxaliplatin to LV5FU2. TRIAL REGISTRATION: NCT00275210 (January 11, 2006). British Journal of Cancer (2019) 121:312–317; https://doi.org/10.1038/s41416-019-0521-7 BACKGROUND provide any survival benefit for low-risk stage II CC patients. Colon cancer (CC) is the third most common cancer in men and Nowadays, there is a need to better identify patients for whom women.1,2 Seventy-five percent of the patients are diagnosed at the addition of oxaliplatin to LV5FU2 can reduce disease relapse a localised stage; where a curative treatment can be proposed. or death, but also to limit useless long-lasting toxicities induced Although adjuvant chemotherapy after curative surgery is by oxaliplatin. recommended for stage III CC patients,3 its survival benefit for Carcinoembryonic antigen (CEA) is a well-known low cost stage II CC is still debated. In this latter group of patients, biological tumour marker used in CC since 1965.6 Previous large adjuvant fluoropyrimidine-based chemotherapy with or without studies have shown that elevated pre-operative CEA levels are oxaliplatin is used. The results of subgroups analyses of large associated with worse prognosis in patients with stage I and II adjuvant trials and meta-analysis have demonstrated several disease, who did not receive adjuvant chemotherapy, and for risk factors for recurrence or death in stage II CC that guide whom pre-operative CEA value of 2.35 ng/mL was defined as an adjuvant treatment decisions.3 In the MOSAIC trial, the optimal cut-off point for survival stratification.7 estimated 10-year probability of overall survival (OS) was The aims of this post hoc study were (i) to assess and validate 75.4% for FOLFOX4 (5-fluorouracil, leucovorin, and oxaliplatin) precisely the prognostic value of post-operative CEA for disease- and 71.7% for LV5FU2 (fluorouracil and leucovorin)4,5 in high- free survival (DFS) and OS in patients with stage II CC treated by risk CC patients with a simplified definition of high-risk stage II adjuvant chemotherapy and (ii) to determine the additional disease (T4, tumour perforation, or <10 examined lymph nodes). predictive value of post-operative CEA for the benefit of the In this study, the addition of oxaliplatin to LV5FU2 did not addition of oxaliplatin to LV5FU2. 1 Department of Hepato-Gastroenterology and Gastrointestinal Oncology, Sorbonne Paris-Cité, Paris Descartes University, Hôpital Européen Georges Pompidou, Paris, France; 2 Methodology and Quality of Life Unit in Oncology, University Hospital of Besançon, Besançon, France; 3Bourgogne Franche-Comté University, INSERM, EFS BFC, UMR1098, Interactions Hôte-Greffon-Tumeur/Ingénierie Cellulaire et Génique, Besançon, France; 4Department of Medical Oncology, Sorbonne University, Hôpital Saint Antoine, Paris, France; 5Oncology Multidisciplinary Research Group (GERCOR), Paris, France; 6UMR-S 1147, INSERM, Paris, France; 7Unit of Medical Oncology, Clinical Cancer Center, AUSL-IRCCS Reggio Emilia, Reggio Emilia, Italy; 8Department of Gastroenterology and Digestive diseases, Hopital Erasme, Université Libre de Bruxelles, Bruxelles, Belgium; 9Vall d’Hebron University Hospital and Institute of Oncology (VHIO), CIBERONC, Universitat Autònoma de Barcelona, Barcelona, Spain; 10Department of Oncology, Royal Bournemouth Hospital and Bournemouth University, Bournemouth, UK and 11Department of Oncology, Institut Hospitalier Franco-Britannique, Levallois-Perret, France Correspondence: Dewi Vernerey (dvernerey@chu-besancon.fr) Received: 5 March 2019 Revised: 12 June 2019 Accepted: 18 June 2019 Published online: 12 July 2019 © The Author(s), under exclusive licence to Cancer Research UK 2019 Association of post-operative CEA with survival and oxaliplatin benefit. . . E Auclin et al. 313 METHODS Table 1. Patient characteristics Population All stage II CC patients from the MOSAIC phase III trial MOSAIC Trial (NCT00275210) were included in this post hoc analysis.4 Post- Stage II CC operative CEA <10 ng/mL was an inclusion criterion in the MOSAIC patients (N = 899) trial. The CEA measurement was not centralised. Age at inclusion, years ≤ 70 798 (88.77%) > 70 101 (11.23%) Definition of high-risk stage II CC Gendera Female 407 (45.27%) High-risk stage II CC was defined as being characterised by one of the following factors: T4, tumour perforation, and <10 or <12 Male 492 (54.73%) examined lymph nodes (the MOSAIC definition and the modified Body mass indexa Underweight 27 (3%) MOSAIC definition, respectively). Normal 438 (48.72%) Overweight 324 (36.04%) Statistical analysis Obese 110 (12.24%) Median values (interquartile range) and frequencies (percentage) Type of adjuvant chemotherapya FOLFOX4 451 (50.17%) were provided for the description of continuous and categorical LV5FU2 448 (49.83%) variables, respectively. Medians and proportions were compared Tumour locationa Left 580 (64.52%) using Student’s t-test and chi-square test (or Fisher’s exact test, if Right 319 (35.48%) appropriate), respectively. Histoprognostic grade G1/2 763 (89.55%) DFS was defined as the time between randomisation and local/ distant relapse, second colorectal/rectal occurrence, or death, G3/4 89 (10.45%) whichever occurred first. Alive patients without relapse and Missing 47 second colorectal/rectal cancer were censored at the date of Stage II risk group Highb 434 (48.65%) their last follow-up. (MOSAIC definition) Lowc 458 (51.35%) OS was defined as the time between randomisation and death Missing 7 1234567890();,: from any cause. Patients known to be alive were censored at the Stage II risk group High 520 (58.3%) date of their last follow-up. (modified MOSAIC definition) Low 372 (41.7%) DFS and OS were estimated using the Kaplan–Meier method Missing 7 and described using median or rate at specific time points with T-stagea T3 728 (80.98%) their 95% confidence intervals (95% CIs). Follow-up was calculated T4 171 (19.02%) using the reverse Kaplan–Meier method. When used continuously, the association between CEA and Number of nodes examineda Median (IQR) 12 (8–19) survival was investigated with the restricted cubic splines method Performance status* 0–1 799 (88.9%) with graphical evaluation. When used as a categorical variable, the ≥2 100 (11.1%) CEA cut-point defined by Margalit et al. in a cohort of 45 449 stage Bowel perforationa 81 (9.01%) I-II CC patients was applied. A sensitivity analysis was performed Bowel obstructiona 158 (17.58%) with the Horton and Lausen method to find the best CEA cut-off Vascular invasion Yes 80 (16.29%) for risk stratification in our cohort.8 Missing 408 The association of demographic, clinical, biological and MMR status dMMR 48 (13.15%) molecular factors with survival was first assessed by the univariate pMMR 317 (86.85%) Cox-proportional-hazard model, providing hazard ratios (HRs) and Missing 534 95% CIs. Parameters with P-values of <0.10 in univariate analysis and/or clinically relevant variables were entered into the multi- CEA level, ng/mL Median (IQR) 1.4 (0.9–2.2) variable Cox-regression model. standard CEA cut-off level, ng/mL ≤5 834 (96.19%) A differential DFS and OS treatment effect among the identified >5 33 (3.81%) CEA risk groups but also when considering the CEA risk groups CEA cut-off level reported by Margalit ≤ 2.35 664 (76.6%) identified as an additional parameter to include in classical well- et al., ng/mL > 2.35 203 (23.4%) established risk group definitions was evaluated with an interaction CEA cut-off level according to Hothorn ≤ 2.77 724 (83.51%) term in the Cox-regression model and illustrated with Kaplan–Meier method, ng/mL > 2.77 143 (16.49%) Missing 32 curves. Time between surgery and CEA Median (IQR) 4.4 (3.3–5.4) All analyses were performed using SAS version 9.4 (SAS measurement, weeks Institute, Cary NC) and R software version 2.15.2 (R Development Follow-up, years Median Median 8.8 (7.9–9.5) Core Team, Vienna, Austria; http://www.r-project.org). P-values of <0.05 were considered statistically significant and a threshold of CC colon cancer, MMR mismatch repair, CEA carcinoembryonic antigen, CI 0.1 was used for interaction terms. All tests were two-sided. confidence interval, IQR interquartile range a no missing data b T4, tumour perforation, or fewer than 10 lymph nodes examined c T1–3 and no tumour perforation and 10 or more lymph nodes examine RESULTS Characteristics of patients Among the 2246 patients included in the MOSAIC trial, 899 had stage II CC. Post-operative CEA was available in 867 (96.4%) patients; 834 (96.2%) had CEA <5 ng/mL and 664 Post-operative CEA and survival (76.6%) <2.35 ng/mL. When post-operative CEA was used as continuous variable, a Overall, 434 (48.65%) and 520 (58.3%) patients had high-risk gradual risk suggesting a linear relation between post-operative stage II CC according to the MOSAIC and the modified MOSAIC CEA and DFS and OS was identified (Fig. 1). Different risk definition, respectively (Table 1). The MSI status was well balanced populations among patients with a post-operative CEA level of between patients with CEA <2.35 and >2.35 ng/mL (12.4% vs <5 ng/mL were determined. This result validates in our cohort the 16.3%, P = 0.36). lower cut-off of 2.35 ng/mL found by Margalit et al. Association of post-operative CEA with survival and oxaliplatin benefit. . . E Auclin et al. 314 a b Hazard ratio for DFS 3 2 Hazard ratio for OS 2 1 1 0.199654 2.090766 4.000980 5.911195 7.802307 0.199654 2.090766 4.000980 5.911195 7.802307 CEA post surgery CEA post surgery Fig. 1 Association between CEA and DFS (a) or OS (b) by the restricted cubic splines method a 1.0 ≤5 ng/mL b 1.0 ≤2.35 ng/mL >5 ng/mL > 2.35 ng/mL Probability of disease free survival Probability of disease free survival 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Logrank p = 0.53 Logrank p = 0.006 0.0 0.0 0 2 4 6 8 10 0 2 4 6 8 10 Time from inclusion (years) Time from inclusion (years) No. at rish No. at rish ≤5 ng/mL 834 739 690 620 343 248 ≤2.35 ng/mL 664 599 563 508 284 203 >5 ng/mL 33 28 23 21 11 9 >2.35 ng/mL 203 168 150 133 70 54 Fig. 2 DFS according to (a) CEA 5 ng/mL and CEA 2.35 ng/mL (b) When post-operative CEA was used as a categorical variable, a Table 2. Multivariate Cox-regression analysis of DFS (N = 860) strong trend was observed between ≤5 and >5 ng/mL values and DFS, with the 3-year DFS rate of 86.6% and 75.5% in the ≤ 5 and > HR 95% CI P 5 ng/mL groups, respectively; P = 0.53. However, the difference was not significant (Fig. 2a). Age, years >70 1.94 1.37–2.76 <0.0001 When using the 2.35 ng/mL cut-off identified by Margalit et al, Gender Male 1.35 1.02–1.79 0.035 patients with a post-operative CEA level ≤2.35 ng/mL were at Tumour location Right 0.78 0.58–1.06 0.111 lower risk of recurrence or death. The 3-year DFS rate was 88.5 and 78.7% for patients with CEA ≤2.35 ng/mL and >2.35 ng/mL, MOSAIC risk group Low 0.74 0.56–0.98 0.035 respectively (P = 0.006; Fig. 2b). Bowel obstruction Yes 1.48 1.07–2.03 0.016 Similar results were observed with OS (Supplementary Fig. S1). In CEA level, ng/mL > 2.35 1.49 1.10–2.00 0.009 the sensitivity analysis, a threshold value of 2.77 ng/mL was identified HR hazard ratio, CI confidence interval, CEA carcinoembryonic antigen as the optimal CEA cut-off (Supplementary Figs. S2 and S3). Independent prognostic value of post-operative CEA In univariate analysis, a post-operative CEA level of >2.35 ng/mL P = 0.009) and OS (HR 1.90, 95% CI 1.33–2.72; P < 0.0001; Table 2 was associated with DFS and OS (DFS: HR 1.51, 95% CI 1.12–2.03; and Supplementary Table S1). P = 0.006 and OS: HR 1.46, 95% CI 1.05–2.05; P = 0.03). After adjustment for age, gender, tumour location, bowel Predictive value of post-operative CEA obstruction, and MOSAIC risk group, CEA >2.35 ng/mL was The benefit of the addition of oxaliplatin to LV5FU2 in term still significantly associated with DFS (HR 1.49, 95% CI 1.10–2.00; of DFS and OS was different between the two CEA groups Association of post-operative CEA with survival and oxaliplatin benefit. . . E Auclin et al. 315 Table 3. Disease-free survival according to treatment arm and CEA level DFS 3 y % (95% CI) P-value for the interaction term between CEA (≤ 2.35, > 2.35) and Treatment Arm (LV5FU2, FOLFOX) All N LV5FU2 FOLFOX LV5FU2 arm FOLFOX arm Absolute Relative HR for treatment arm N arm N changed changee effectc (95% CI) Whole population 899 448 451 84.7 (81.4–88.1) 87.1 (84–90) +2.4 +2.8 0.11 CEA ≤2.35 664 333 331 88.2 (84.8–91.7) 88.7 (85.4–92.2) +0.5 +0.6 1.07 (0.77–1.47) CEA >2.35 203 97 106 76 (67.9–85.1) 81.1 (74–88.9) +5.1 +6.7 0.67 (0.41–1.11) High-riska population 434 222 212 81.3 (76.2–86.6) 86.3 (81.7–91) +5 +6.1 0.09 High-risk - CEA ≤2.35 320 158 162 84 (78.4–89.9) 86.3 (81.2–91.8) +2.3 +2.7 0.95 (0.62–1.45) High-risk - CEA >2.35 102 55 47 74 (63.2–86.7) 87.2 (78.2–97.3 +13.2 +17.8 0.47 (0.23–0.97) Low-riskb population 458 223 235 87.9 (83.7–92.3) 88 (83.9–92.3) +0.1 +0.1 0.78 Low-risk - CEA ≤2.35 338 172 166 91.8 (87.8–96) 91.5 (87.3–95.8) −0.3 −0.3 1.18 (0.71–1.95) Low-risk - CEA >2.35 100 42 58 78.6 (67.1–92) 75.8 (65.6–87.7) −2.8 −3.6 0.99 (0.47–2.10) Absolute difference at time X gives the difference between percentages observed in the two treatment arms at time X; the relative difference at time X gives the proportion of increase or decrease in survival rate of one arm relative to the other arm at time X a T4, tumour perforation, or fewer than 10 lymph nodes examined b T1–3 and no tumour perforation and 10 or more lymph nodes examined c HR for treatment effect (the addition of oxaliplatin to the LV5FU2 regimen) d Absolute difference reflects a comparison of survival between the FOLFOX and LV5FU2 arms e Relative difference reflects a ratio of the observed survival in the FOLFOX arm and the LV5FU2 arm [(X year OS rate in the FOLFOX group - X year OS rate in the LV5FU2 group)/(X year OS rate in the LV5FU2 group) *100] a 1.0 b 1.0 + ox c – ox 1.0 Probability of disease free survival Probability of disease free survival Probability of disease free survival 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 CEA ≤2.35 without ox CEA ≤2.35 without ox CEA ≤2.35 with ox CEA ≤2.35 with ox CEA >2.35 without ox CEA >2.35 without ox Logrank p = 0.01 Logrank p = 0.90 Logrank p = 0.04 0.0 CEA >2.35 with ox 0.0 CEA >2.35 with ox 0.0 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 Time from inclusion (years) Time from inclusion (years) Time from inclusion (years) No. at risk No. at risk No. at risk CEA ≤2.35 without ox 333 300 281 258 141 100 + ox 235 212 199 175 97 64 CEA ≤2.35 without ox 172 159 154 145 81 57 CEA ≤2.35 with ox 331 299 282 250 143 103 – ox 223 198 190 177 97 73 CEA ≤2.35 with ox 166 153 146 127 72 50 CEA >2.35 without ox 97 75 67 60 27 24 CEA >2.35 without ox 42 33 32 28 13 13 CEA >2.35 with ox 106 93 83 73 43 30 CEA >2.35 with ox 58 48 43 38 19 10 d 1.0 + ox e 1.0 – ox Probability of disease free survival Probability of disease free survival 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 CEA ≤2.35 without ox CEA ≤2.35 with ox Logrank p = 0.19 Logrank p = 0.07 CEA >2.35 without ox 0.0 0.0 CEA >2.35 with ox 0 2 4 6 8 10 0 2 4 6 8 10 Time from inclusion (years) Time from inclusion (years) No. at risk No. at risk + ox 212 191 174 157 95 73 CEA ≤2.35 without ox 158 138 124 110 58 41 – ox 222 188 166 149 77 56 CEA ≤2.35 with ox 162 144 134 122 71 53 CEA >2.35 without ox 55 42 35 32 14 11 CEA >2.35 with ox 47 44 39 34 23 19 Fig. 3 Benefit for DFS with the addition of oxaliplatin to LV5FU2 in a all patients, b the MOSAIC low-risk patients, c the MOSAIC low-risk patients after CEA stratification, d the MOSAIC high-risk patients, and e the MOSAIC high-risk patients after CEA stratification. Figure 3 abbreviations: Ox oxaliplatin; +ox with oxaliplatin; Ox without oxaliplatin (Table 3 and Supplementary Table S2). Only patients with high Among patients with high post-operative CEA levels, the benefit post-operative CEA levels seemed to have survival benefit from from oxaliplatin addition was observed in the high-risk stage II CC oxaliplatin addition to LV5FU2 (interaction terms P = 0.09; patients (DFS interaction term P = 0.09 and OS interaction term Fig. 3a). P = 0.03), but not in those with low-risk stage tumours (DFS Association of post-operative CEA with survival and oxaliplatin benefit. . . E Auclin et al. 316 a 1.0 b 1.0 + ox c 1.0 – ox Probability of overall survival Probability of overall survival Probability of overall survival 0.8 0.8 0.8 0.6 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.2 CEA ≤2.35 without ox CEA ≤2.35 without ox CEA ≤2.35 with ox CEA ≤2.35 with ox Logrank p = 0.01 CEA >2.35 without ox Logrank p = 0.51 Logrank p = 0.51 CEA >2.35 without ox 0.0 CEA >2.35 with ox 0.0 0.0 CEA >2.35 with ox 0 2 4 6 8 10 0 2 4 6 8 10 0 2 4 6 8 10 Time from inclusion (years) Time from inclusion (years) Time from inclusion (years) No. at risk No. at risk No. at risk CEA ≤2.35 without ox 333 323 308 283 154 106 + ox 235 226 215 188 101 68 CEA ≤2.35 without ox 172 170 162 153 85 60 CEA ≤2.35 with ox 331 316 298 265 146 106 – ox 223 216 203 190 104 77 CEA >2.35 with ox 166 159 153 133 72 51 CEA >2.35 without ox 97 86 77 68 30 25 CEA >2.35 without ox 42 38 35 32 15 14 CEA >2.35 with ox 106 102 95 82 47 33 CEA ≤2.35 with ox 58 56 52 45 23 13 d 1.0 + ox e 1.0 – ox Probability of overall survival Probability of overall survival 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 CEA ≤2.35 without ox CEA ≤2.35 with ox Logrank p = 0.58 Logrank p = 0.01 CEA >2.35 without ox 0.0 0.0 CEA >2.35 with ox 0 2 4 6 8 10 0 2 4 6 8 10 Time from inclusion (years) Time from inclusion (years) No. at risk No. at risk + ox 212 203 186 168 98 75 CEA ≤2.35 without ox 158 150 143 127 67 44 – ox 222 206 193 171 87 59 CEA ≤2.35 with ox 162 155 143 131 74 55 CEA >2.35 without ox 55 48 42 36 15 11 CEA >2.35 with ox 47 45 42 36 23 19 Fig. 4 Benefit for OS with the addition of oxaliplatin to LV5FU2 in (a) all patients, (b) the MOSAIC low-risk patients, (c) the MOSAIC low-risk patients after CEA stratification, (d) the MOSAIC high-risk patients, and (e) the MOSAIC high-risk patients after CEA stratification. Figure 4 abbreviations: Ox: oxaliplatin; + ox: with oxaliplatin; Ox: without oxaliplatin interaction term P = 0.78 and OS interaction term P = 1). Patients those who had CEA <2.35 ng/mL in multivariate analysis. We with high-risk stage II tumours and CEA >2.35 ng/mL represent analysed the post-operative CEA level given the results by Konishi 25% of all stage II patients included in this post hoc analysis. The et al. who showed that the pre-operative CEA level is not a addition of oxaliplatin had no survival advantage in patients with relevant marker of recurrence in localised CC if CEA is normalised low-risk stage II disease and CEA < 2.35 ng/mL (75% of the study after surgery.10 population; Fig. 3c–e; Table 3 and Supplementary Table S2). We showed in our study that only patients with high-risk stage II These results were replicated for the OS endpoint (Fig. 4) and CC and high post-operative CEA level benefited from the addition using the modified MOSAIC definition for risk groups (Supple- of oxaliplatin in terms of DFS, with a 13% absolute increase in mentary Figs. S4 and S5). When assessed in the population aged 3-year DFS rate. Another predictive factor of oxaliplatin benefit ≤70 years, the predictive value of CEA level was also observed in that has been published, is a high-risk score with the Oncotype Dx the high-risk group (Supplementary Figs. S6 and S7). genomic signature.11 In fact, Yothers et al. showed, in a limited Similar results were observed with the optimal CEA cut-off value cohort of patients, that only stage II CC with high risk Oncotype Dx of 2.77 ng/mL (Supplementary Table S3). genomic signature seemed to benefit from the addition of oxaliplatin to adjuvant therapy. Circulating tumour DNA (ctDNA) has shown remarkable DISCUSSION prognostic value for OS and RFS in very small cohorts of patients In this post hoc analysis of the MOSAIC study, post-operative with early-stage CC.12–14 Lu et al. suggested that the combination serum CEA level, with the cut-off of 2.35 ng/mL, was an of post-operative serum CEA levels assessment and persistent independent prognostic factor in CC patients with stage II disease. post-operative circulating tumour cells detection is prognostic Moreover, CEA appeared to be a predictive factor of the benefit predictor of early relapse in stage II–III CC patients.15 from the addition of oxaliplatin to LV5FU2 in adjuvant therapy for All the above results suggest that a minimal residual disease high-risk stage II CC. may be associated with ctDNA, circulating tumour cells, CEA, or The prognostic value of post-operative CEA has been demon- prognostic genomic signatures and that patients harbouring these strated for stage I to IV CC.9 Our results are in line with the factors are those who can benefit most from treatment literature, as we also showed that a lower cut-off of CEA is more intensification. suitable level for death or recurrence risk stratification. The cut-off In our study, we used the risk groups definition published by values of 2.35 ng/mL (Margalit et al. study)7 and 2.77 ng/mL André et al in 2015 (T4, tumour perforation, or less than 10 (defined in our cohort study) suggest that the common 5 ng/mL examined lymph nodes). This definition was published in the cut-off is clearly not optimal for risk stratification of patients with MOSAIC population, in order to lose as less information as CC. The magnitude of the prognostic value of the CEA level was possible. As it is slightly different from the ESMO consensus important in our study. Indeed, patients with high level of CEA definition (12 lymph nodes examined instead of 10), we also had a 50% increased risk of death or recurrence compared to validated our results with the latter definition (modified MOSAIC Association of post-operative CEA with survival and oxaliplatin benefit. . . E Auclin et al. 317 definition). Moreover, in order to deal with potential confounding Ethics approval and consent to participate: The MOSAIC trial was performed in factors related to comorbidities, we performed sensitivity analysis accordance with the Declaration of Helsinki. All patients have given their written in the ≤70 years’ population. Again, similar results were observed, consent to be enroled in the MOSAIC trial. Written informed consent was required adding to their reliability. from all patients and the study was approved by the ethics committees of the participating centres. The use of the data for translational research was approved by In addition, in patients with high-risk stage II CC and post- the ethic committee of Ile De France (April 29, 2009). operative CEA >2.35 ng/mL most deaths occurred within 2 years after surgery. This strongly suggest that post-operative imaging Funding: None. including PET-scan should be performed in those patients in order to further look at metastases before engaging adjuvant therapy, especially in randomised clinical trials. Consent to publish: Not applicable. Our results are robust and obtained with the best material available to study the interest to consider post-operative CEA Data availability: Dataset of the study can be found with the corresponding author. information among classical high and low-risk classification in order to better identify patients who can benefit most from the addition Note: This work is published under the standard license to publish agreement. After of oxaliplatin to LV5FU2. Ideally, our findings should be validated in 12 months the work will become freely available and the license terms will switch to an external cohort data. Unfortunately, such data do not exist. a Creative Commons Attribution 4.0 International (CC BY 4.0). Indeed, no randomised trial that have studied the addition of oxaliplatin to LV5FU2 in stage II CC patients provide precise post- Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims operative CEA data. Another limit of this study is the lack of in published maps and institutional affiliations. smoking habit data, a possible confusion factor. Therefore, clinician should interpret these results with caution in smoking patients. REFERENCES 1. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2018. CA. Cancer J. Clin. 68, 7–30 (2018). CONCLUSION 2. Malvezzi, M., Bertuccio, P., Levi, F., La Vecchia, C. & Negri, E. European cancer Our results show that post-operative CEA is a strong prognostic mortality predictions for the year 2014. Ann. Oncol. 25, 1650–1656 (2014). factor for DFS and OS in stage II CC. In the MOSAIC trial, only high- 3. Labianca, R., Nordlinger, B., Beretta, G. D., Mosconi, S., Mandalà, M., Cervantes, A. risk stage II CC with post-operative CEA >2.35 ng/mL (~25% of our et al. Early colon cancer: ESMO clinical practice guidelines for diagnosis, treat- stage II population) were identified to benefit from the addition of ment and follow-up. Ann. Oncol. 24(Suppl 6), vi64–72 (2013). oxaliplatin to LV5FU2. CEA >2.35 ng/mL should be included in the 4. André, T., Boni, C., Mounedji-Boudiaf, L., Navarro, M., Tabernero, J., Hickish, T. et al. definition of high-risk stage II CC and should lead to post-operative Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. check-up before adjuvant therapy. Moreover, it should be included N. Engl. J. Med. 350, 2343–2351 (2004). 5. André, T., de Gramont, A., Vernerey, D., Chibaudel, B., Bonnetain, F., Tijeras- in the future trials assessing adjuvant strategies in stage II CC. Raballand, A. et al. Adjuvant fluorouracil, leucovorin, and oxaliplatin in stage II to III Colon cancer: updated 10-year survival and outcomes according to BRAF mutation and mismatch repair status of the MOSAIC study. J. Clin. Oncol. 33, ACKNOWLEDGEMENTS 4176–4187 (2015). The editorial assistance with preparation of the manuscript was provided by 6. Gold, P. & Freedman, S. O. Specific carcinoembryonic antigens of the human Magdalena Benetkiewicz (PhD, GERCOR). digestive system. J. Exp. Med. 122, 467–481 (1965). 7. Margalit, O., Mamtani, R., Yang, Y.-X., Reiss, K. A., Golan, T., Halpern, N. et al. Assessing the prognostic value of carcinoembryonic antigen levels in stage I and AUTHOR CONTRIBUTIONS II colon cancer. Eur. J. Cancer. 94, 1–5 (2018). E.A.: data analysis, manuscript writing. T.A.: data collection, manuscript writing and 8. Hothorn, T. & Lausen, B. On the exact distribution of maximally selected rank reviewing. J.T.: manuscript writing and reviewing. M.B., J.L.V.L., J.T., T.H. and A.G.: data statistics. Comput. Stat. Data. Anal. 43, 121–137 (2003). collection, manuscript reviewing. D.V.: data analysis, manuscript writing and reviewing. 9. Thirunavukarasu, P., Sukumar, S., Sathaiah, M., Mahan, M., Pragatheeshwar, K. D., Pingpank, J. F. et al. C-stage in colon cancer: implications of carcinoembryonic antigen biomarker in staging, prognosis, and management. J. Natl. Cancer. Inst. 103, 689–697 (2011). ADDITIONAL INFORMATION 10. Konishi, T., Shimada, Y., Hsu, M., Tufts, L., Jimenez-Rodriguez, R., Cercek, A. et al. Supplementary information is available for this paper at https://doi.org/10.1038/ Association of preoperative and postoperative serum carcinoembryonic antigen s41416-019-0521-7. and colon cancer outcome. JAMA Oncol. 4, 309–315 (2018). 11. Yothers, G., O’Connell, M. J., Lee, M., Lopatin, M., Clark-Langone, K. M., Millward, C. Competing interests: Thierry André: Honoraria—AMGEN; Bayer; Bristol-Myers et al. Validation of the 12-gene colon cancer recurrence score in NSABP C-07 as a Squibb; PRMA Consulting; Roche/Genentech; Sanofi; SERVIER; Xbiotech. Consulting predictor of recurrence in patients with stage II and III colon cancer treated with or Advisory Role—Amgen; Bayer; Bristol-Myers Squibb; Guardant Health; HalioDX; fluorouracil and leucovorin (FU/LV) and FU/LV plus oxaliplatin. J. Clin. Oncol. 31, MSD Oncology; Mundipharma; Roche/Genentech; SERVIER. Travel, Accommodations, 4512–4519 (2013). Expenses—Amgen; Bristol-Myers Squibb; MSD Oncology; Roche/Genentech. Julien 12. Fan, G., Zhang, K., Yang, X., Ding, J., Wang, Z. & Li, J. Prognostic value of circu- Taieb: Consulting or Advisory Role—Amgen; Baxalta; Celgene; Lilly; Merck KGaA; lating tumor DNA in patients with colon cancer: systematic review. PLoS ONE. 12, Roche; SERVIER; Sirtex Medical. Speakers’ Bureau—Amgen; Baxalta; Lilly; Merck; e0171991 (2017). Roche/Genentech; Sanofi; SERVIER. Maria Banzi: Honoraria—Merck; Roche (I). Travel, 13. Lecomte, T., Berger, A., Zinzindohoué, F., Micard, S., Landi, B., Blons, H. et al. Accommodations, Expenses—Merck; Pfizer; Roche. Josep Tabernero: Consulting or Detection of free-circulating tumor-associated DNA in plasma of colorectal can- Advisory Role—Amgen; Boehringer Ingelheim; Celgene; Chugai Pharma; ImClone cer patients and its association with prognosis. Int. J. Cancer 100, 542–548 (2002). Systems; Lilly; Merck; Merck Serono; Millennium; Novartis; Roche/Genentech; Sanofi; 14. Tie, J., Wang, Y., Tomasetti, C., Li, L., Springer, S., Kinde, I. et al. Circulating tumor Taiho Pharmaceutical. Tamas Hickish: Stock and Other Ownership Interests— DNA analysis detects minimal residual disease and predicts recurrence in patients Knowledge To Practice. Patents, Royalties, Other Intellectual Property—Bournemouth with stage II colon cancer. Sci. Transl. Med. 8, 346ra92 (2016). University (Inst). Travel, Accommodations, Expenses—Philips Healthcare; XBiotech. 15. Lu, C.-Y., Uen, Y.-H., Tsai, H.-L., Chuang, S.-C., Hou, M.-F., Wu, D.-C. et al. Molecular Aimery De Gramont: Honoraria—Chugai Pharma; Yakult. Dewi Vernerey: Consulting detection of persistent postoperative circulating tumour cells in stages II and III or Advisory Role—HalioDx; Janssen-Cilag; OSE Immunotherapeutics; Prestizia. The colon cancer patients via multiple blood sampling: prognostic significance of remaining authors declare no competing interests. detection for early relapse. Br. J. Cancer 104, 1178–1184 (2011).