Abstract
Background:
The prognostic value of KRAS and BRAFV600E mutations in stage III colon cancer (CC) remains controversial and has never been clearly analyzed in patients with microsatellite instability-high (MSI-H) tumors due to sample size limitations. Data are also lacking for KRAS submutations and prognosis.
Patients and methods:
We examined clinicopathological variables and prognosis in patients with surgically resected stage III CC who participated in seven clinical trials from the ACCENT/IDEA databases. Associations between KRAS exon 2 and BRAFV600E mutations and time to recurrence (TTR), overall survival (OS), and survival after recurrence (SAR) were assessed using a Cox model. We also analyzed the prognostic value of KRAS exon 2 submutations.
Results:
Among 8460 patients, 11.4% had MSI-H status. In the MSI-H group, BRAFV600E, KRAS exon 2 mutants, and double-wild-type statuses were detected in 40.6%, 18.1%, and 41.3%, respectively, whereas and in the microsatellite stable (MSS) group, these were detected in 7.7%, 38.6%, and 53.8%, respectively. In the MSS group, 5-year TTR rates of 61.8%, 66.3%, and 72.9% were observed among patients with BRAFV600E, KRAS exon 2 mutants, and those who were DWT, respectively [adjusted hazard ratio (HR) = 1.58 and 1.31, both P < 0.001]. In the MSI-H group, 5-year TTR rates did not differ significantly among the mutated subgroups. Similar results were found for OS. However, survival after relapse was significantly shorter in the KRAS exon 2- and BRAFV600E-mutated patients in both MSS (adjusted HR = 2.06 and 1.15; both P < 0.05) and MSI-H (adjusted HR = 1.99 and 1.81; both P < 0.05) groups. In the MSS group, KRAS exon 2 mutations were associated with TTR, but only p.G12C, p.G12D, and p.G13D were associated with poor outcomes after disease recurrence.
Conclusions:
Testing for both KRAS and BRAFV600E mutations in stage III patients should be considered as they can better define individual patient prognosis, and may also enable patient selection for (neo)adjuvant trials dedicated to specific molecular subtypes with poor prognosis.
Keywords: stage III colon cancer, RAS mutation, BRAFV600E mutation, MSI/dMMR, adjuvant, survival after relapse, time to recurrence
INTRODUCTION
The management of colon cancer (CC) has been radically transformed over the past 20 years.1,2 Indeed, molecular profiling of these tumors in the metastatic setting has shown that CC is heterogeneous and composed of distinct subgroups with varying clinical and biological characteristics, and different prognoses and therapeutic options.3
International guidelines currently recommend routine testing of all CCs for microsatellite instability (MSI) and the mutational status of the tumor for KRAS, NRAS, and BRAF genes particularly in the metastatic setting.2,4 The emergence of specific therapeutic options for MSI-H and BRAFV600E-mutated CC, validated by large phase III studies, has translated to the individualization of therapies for these patients.5,6 In addition, the presence of RAS mutations indicates tumor resistance to epidermal growth factor receptor inhibitors and precludes treating those patients with those monoclonal antibodies. In addition, the emergence of promising treatments for other molecular subgroups of CC such as the subgroup with both MSI-H status and a BRAFV600E mutation, a KRAS p.G12C mutation,7 an HER2 gene amplification,8,9 or a NTRK gene fusion10 suggests that the molecular segmentation, of what was once considered as a single disease, will increase in the forthcoming years.
In patients with localized disease, the determination of the MSI status of the tumor is currently recommended, as it brings prognostic information, allows screening for Lynch syndrome, and may lead to the early use of immuno-oncology agents. Other genomic abnormalities frequently investigated in the metastatic setting are currently not commonly assessed at earlier disease stages.11 However, therapeutic successes observed with targeted therapies in the metastatic setting have led to the conduct of clinical studies of targeted agents and immunotherapy in nonmetastatic patients in the adjuvant or neoadjuvant settings with promising results. Knowledge of the tumor molecular profile in patients with stage III CC is increasingly important to accurately design clinical trials and to determine the relevant statistical assumptions and stratification factors for these studies.
ACCENT and IDEA are two international consortia dedicated to pooled analyses of CC clinical trials in the adjuvant setting. In this article, we analyze all patients with resected stage III CC from the ACCENT and IDEA databases for whom molecular profile data were available. Our goals were to analyze the prognosis linked to KRAS exon 2 and BRAFV600E mutations in patients with stage III microsatellite stable (MSS) tumors and to analyze with a significant number of patients their prognostic value in those with MSI-H tumors, together with the impact of KRAS exon 2 submutations.
PATIENTS AND METHODS
Patients
All patients with stage III CC who participated in any of the seven relevant randomized clinical trials of adjuvant chemotherapy whose data are in the ACCENT and IDEA databases who had available MSI, BRAF, and KRAS status results were included.
The following trials were included in the analyses: IDEA France (NCT00958737), NSABP-C07 (NCT00004931) and C08 (NCT00096278), C89803 (NCT00003835), PETACC3 (NCT00026273), PETACC8 (NCT00265811), and NCCTG-N0147 (NCT00079274). Treatment arms in these studies were fluoropyrimidine alone; FOLFOX + oxaliplatin and capecitabine (CAPOX; also called XELOX); FOLFOX + cetuximab; FOLFOX + bevacizumab; leucovorin calcium (folinic acid), fluorouracil, and irinotecan hydrochloride (FOLFIRI); or FOLFIRI + cetuximab. We excluded patients with stage I, II, and IV CC, or lower-middle rectal cancer, or who never received chemotherapy.
This is a secondary analysis of previously conducted, pooled clinical trial data, and was approved by the Mayo Clinic Institutional Review Board. Patients provided informed consent prior to participating in the original trials, which was obtained by local enrolling centers. Written informed consent was required from each patient at protocol entry to permit the use of their tissue for molecular profiling in preplanned translational research.
Molecular profiling
Mismatch repair (MMR) tumor status was specified in each study and was determined by immunohistochemistry (IHC) or by PCR MSI testing when IHC was indeterminate, consistent with current recommendations. MSI-H tumors were defined as those that exhibited either the loss of expression of one or more MMR proteins by IHC or were classified as having high-level MSI (MSI-H) on PCR testing. MSS tumors were defined by normal MMR protein expression by IHC or were classified as MSS or low-level MSI status based on MSI testing.
Supplementary Table S1, available at https://doi.org/10.1016/j.annonc.2023.08.006, summarizes how KRAS, BRAF, and MSI statuses were assessed in the different trials included in this pooled analysis.
Statistical analyses
The outcomes in the overall population were considered and then the subgroups comprising patients with MSS and MSI-H were separately considered for the reported analysis. Patients were divided into three groups. Group 1: wild type for both KRAS exon 2 and BRAFV600E (double-wild type, designated as DWT); group 2: mutant for KRAS exon 2 (KRAS MT); and group 3: mutant for BRAFV600E (BRAF MT). As both other groups exhibit poor prognosis in the metastatic setting, we chose to use the DWT group as our reference group for all our analyses. Comparisons of patients with specific codon 12 and codon 13 KRAS mutations versus the DWT population were also performed. Codon 12 p.G12A, p.G12R, p.G12S, and other rare codon 12 mutations were grouped as ‘KRAS MT other’ due to the small numbers of tumors in each of these subgroups. By contrast, outcomes in tumors exhibiting KRAS mutants p.G12V, p.G12D, p.G12C, and p.G13D were analyzed individually.
The endpoints for these analyses were time to recurrence (TTR), overall survival (OS), and survival after recurrence (SAR). TTR was defined as the time from randomization to recurrence or last disease evaluation. OS was defined as the time from randomization to death due to any cause or last known date alive. SAR was defined as the time from disease recurrence to death due to any cause or last known date alive.
TTR, OS, and SAR curves were estimated with the Kaplan—Meier method. Differences between groups of patients were analyzed using stratified log-rank tests and a stratified Cox model. To control for potential confounding effects, the main Cox proportional hazard models were built by adjusting for known prognostic factors, namely, age, sex, World Health Organization (WHO) performance status (PS), tumor stages pT and pN, tumor grade, and tumor sidedness, and were then stratified by clinical trial. Analyses were done with SAS (version 9.4; SAS Institute, Cary, NC). Two-sided P values <0.05 were considered statistically significant and were not adjusted for multiple comparisons.
RESULTS
Study population
Among the 13 470 patients included in the seven phase III studies, 11 850 patients were in stage III, 9299 had data on KRAS and BRAFV600E tumors, and 8460 also had MSS/MSI status data (Figure 1). Demographic and clinical characteristics of the patients in the molecular study (n = 8460) were not significantly different from those in the randomized population (n = 11 850; Supplementary Table S2, available at https://doi.org/10.1016/j.annonc.2023.08.006).
Figure 1. Study flowchart.
BRAF MT, mutant for BRAFV600E; KRAS MT, mutant for KRAS exon 2; MSI, microsatellite instability; MSS, microsatellite stable; WT, wild type.
KRAS and BRAFV600E statuses and clinicopathological variables
In the overall study population, 8460 patients had available molecular data, of which 52.4% were DWT, 36.2% had a KRAS-mutant exon 2, and 11.4% had a BRAFV600E mutation. KRAS mutations were located on codon 12 in 26.6% of the cases and on codon 13 in 8.0% of the cases. Clinical and pathological characteristics associated with KRAS MT and BRAF MT are summarized in Supplementary Table S3, available at https://doi.org/10.1016/j.annonc.2023.08.006.
In line with literature, an MSI-H tumor was found in 11.4% of the population, while the majority was MSS (88.6%).
Among the 7492 MSS tumors, 53.8% were DWT, 38.6% KRAS mutant, and 7.7% BRAFV600E mutant (Figure 1). Those patients with a BRAFV600E mutation in their tumors were older, more likely female, of white race, had higher T and N stage, poorly differentiated histology, and tumors located in the proximal colon compared with DWT tumors (Table 1). Patients with KRAS-mutant tumors were more likely female and had a tumor originating in a proximal location compared with those with DWT tumors (Table 1).
Table 1.
Baseline patient characteristics
| Characteristics | MSS/MSI-low/pMMR | MSI-high/dMMR | ||||||
|---|---|---|---|---|---|---|---|---|
| DWT (N = 4029) |
KRAS MT (N = 2889) |
BRAF MT (N = 574) |
P value | DWT (N = 400) |
KRAS MT (N = 175) |
BRAF MT (N = 393) |
P value | |
| Age | <0.0001* | <0.0001* | ||||||
| Mean (SD) | 58.1 (10.68) | 59.8 (10.17) | 62.3 (9.74) | 54.0 (12.51) | 49.9 (12.28) | 66.3 (8.12) | ||
| Median (IQR) | 59.0 (51.0-66.0) | 61.0 (53.0- 67.0) | 63.0 (55.0-70.0) | 54.0 (45.0-64.0) | 50.0 (41.0-57.0) | 67.0 (62.0-72.0) | ||
| Age group, n (%) | <0.0001** | <0.0001** | ||||||
| <70 years | 3560 (88.4) | 2479 (85.8) | 449 (78.2) | 359 (89.8) | 162 (92.6) | 273 (69.5) | ||
| ≥70 years | 469 (11.6) | 410 (14.2) | 125 (21.8) | 41 (10.3) | 13 (7.4) | 120 (30.5) | ||
| Sex, n (%) | <0.0001** | <0.0001** | ||||||
| Male | 2377 (59.0) | 1547 (53.5) | 283 (49.3) | 231 (57.8) | 113 (64.6) | 111 (28.2) | ||
| Female | 1652 (41.0) | 1342 (46.5) | 291 (50.7) | 169 (42.3) | 62 (35.4) | 282 (71.8) | ||
| Race | <0.0001** | 0.1019** | ||||||
| White, n (%) | 2384 (88.6) | 1663 (87.8) | 388 (95.3) | 259 (91.5) | 117 (87.3) | 284 (92.8) | ||
| Black, n (%) | 149 (5.5) | 150 (7.9) | 6 (1.5) | 12 (4.2) | 7 (5.2) | 16 (5.2) | ||
| Asian, n (%) | 106 (3.9) | 51 (2.7) | 7 (1.7) | 5 (1.8) | 4 (3.0) | 5 (1.6) | ||
| Other, n (%) | 51 (1.9) | 29 (1.5) | 6 (1.5) | 7 (2.5) | 6 (4.5) | 1 (0.3) | ||
| Missing, n | 1339 | 996 | 167 | 117 | 41 | 87 | ||
| Performance score, n (%) | 0.0078** | 0.0026** | ||||||
| 0, n (%) | 3180 (79.5) | 2236 (78.0) | 420 (73.9) | 316 (79.6) | 141 (82.0) | 277 (70.8) | ||
| 1+, n (%) | 820 (20.5) | 629 (22.0) | 148 (26.1) | 81 (20.4) | 31 (18.0) | 114 (29.2) | ||
| Missing, n | 29 | 24 | 6 | 3 | 3 | 2 | ||
| BMI (kg/m2) | 0.1151* | 0.0004* | ||||||
| Mean (SD) | 27.3 (5.65) | 27.0 (5.56) | 26.9 (4.87) | 26.1 (5.33) | 26.5 (5.81) | 27.5 (5.62) | ||
| Median (IQR) | 26.5 (23.4–30.1) | 26.1 (23.3–29.9) | 26.4 (23.6–29.7) | 25.1 (22.7–28.5) | 25.9 (22.3–29.1) | 26.9 (23.3–30.1) | ||
| T-stage, n (%) | <0.0001** | 0.2958** | ||||||
| T1 | 147 (3.6) | 75 (2.6) | 18 (3.1) | 5 (1.3) | 4 (2.3) | 1 (0.3) | ||
| T2 | 439 (10.9) | 277 (9.6) | 30 (5.2) | 29 (7.3) | 11 (6.3) | 24 (6.1) | ||
| T3 | 2945 (73.1) | 2118 (73.3) | 428 (74.6) | 293 (73.3) | 126 (72.0) | 305 (77.6) | ||
| T4 | 498 (12.4) | 419 (14.5) | 98 (17.1) | 73 (18.3) | 34 (19.4) | 63 (16.0) | ||
| N-stage (calculated), n (%) | <0.0001** | 0.8826** | ||||||
| N1 | 2557 (63.5) | 1911 (66.1) | 284 (49.5) | 252 (63.0) | 111 (63.4) | 242 (61.6) | ||
| N2 | 1472 (36.5) | 978 (33.9) | 290 (50.5) | 148 (37.0) | 64 (36.6) | 151 (38.4) | ||
| Risk group, n (%) | <0.0001** | 0.6570** | ||||||
| Stage III: T1-3N1 | 2280 (56.6) | 1658 (57.4) | 237 (41.3) | 217 (54.3) | 88 (50.3) | 205 (52.2) | ||
| Stage III: T4 or N2 | 1749 (43.4) | 1231 (42.6) | 337 (58.7) | 183 (45.8) | 87 (49.7) | 188 (47.8) | ||
| Grouped differential grade | <0.0001** | 0.0003** | ||||||
| Grade I-II, n (%) | 3366 (84.0) | 2422 (84.7) | 360 (63.9) | 214 (54.6) | 113 (66.5) | 187 (47.9) | ||
| Grade III-IV, n (%) | 639 (16.0) | 438 (15.3) | 203 (36.1) | 178 (45.4) | 57 (33.5) | 203 (52.1) | ||
| Missing, n | 24 | 29 | 11 | 8 | 5 | 3 | ||
| Sidedness | <0.0001** | <0.0001** | ||||||
| Proximal, n (%) | 1196 (30.1) | 1413 (50.1) | 400 (70.9) | 299 (77.1) | 119 (70.4) | 360 (92.8) | ||
| Distal, n (%) | 2752 (69.3) | 1378 (48.9) | 158 (28.0) | 80 (20.6) | 47 (27.8) | 26 (6.7) | ||
| Both, n (%) | 24 (0.6) | 27 (1.0) | 6 (1.1) | 9 (2.3) | 3 (1.8) | 2 (0.5) | ||
| Missing, n | 57 | 71 | 10 | 12 | 6 | 5 | ||
| Treatment, n (%) | <0.0001** | 0.0008** | ||||||
| 5FU | 549 (13.6) | 392 (13.6) | 75 (13.1) | 73 (18.3) | 39 (22.3) | 57 (14.5) | ||
| FOLFOX | 1848 (45.9) | 1565 (54.2) | 260 (45.3) | 163 (40.8) | 81 (46.3) | 162 (41.2) | ||
| FOLFIRI | 385 (9.6) | 259 (9.0) | 65 (11.3) | 48 (12.0) | 29 (16.6) | 48 (12.2) | ||
| FOLFOX + cetuximab | 1123 (27.9) | 575 (19.9) | 149 (26.0) | 108 (27.0) | 20 (11.4) | 114 (29.0) | ||
| FOLFIRI + cetuximab | 16 (0.4) | 11 (0.4) | 4 (0.7) | 1 (0.3) | 3 (1.7) | 4 (1.0) | ||
| FOLFOX + bevacizumab | 68 (1.7) | 47 (1.6) | 11 (1.9) | 2 (0.5) | 1 (0.6) | 7 (1.8) | ||
| CAPOX (XELOX) | 40 (1.0) | 40 (1.4) | 10 (1.7) | 5 (1.3) | 2 (1.1) | 1 (0.3) | ||
5-FU, fluorouracil; BMI, body mass index; CAPOX/XELOX, oxaliplatin and capecitabine; dMMR, deficient mismatch repair; DWT, double wild type; FOLFIRI, leucovorin calcium (folinic acid), fluorouracil, and irinotecan hydrochloride; IQR, interquartile range; MSI, microsatellite instability; MSS, microsatellite-stable; MT, mutated; pMMR, mismatch repair proficient; SD, standard deviation.
Kruskal—Wallis P value.
Chi-square P value.
Among the 968 MSI-H tumors, 41.3% were DWT, 18.1% KRAS mutant, and 40.6% BRAFV600E mutant (Figure 1). For patients with MSI-H tumors, those with BRAFV600E mutations were older, more likely female, had higher grade histology, and tumors located in the proximal colon. Tumors with mutant KRAS occurred more commonly in males and had lower grade histology (Table 1) compared with DWT MSI-H tumors.
Outcome predictors in the overall study population
The median follow-up on the pooled population was 6.9 years [95% confidence interval (CI) 6.8-6.9 years], 6.9 years (95% CI 6.9-7.0 years), and 7.0 years (95% CI 6.9-7.1 years) for patients with DWT tumors, KRAS-mutated tumors, and BRAFV600E-mutated tumors, respectively. Both mutations were associated with shorter TTR, OS, and SAR compared with DWT patients (Supplementary Figure S1, available at https://doi.org/10.1016/j.annonc.2023.08.006). In a multi-variate analysis including all clinically or statistically significant variables, both mutations remained associated with shorter TTR, OS, and SAR (Supplementary Table S4, available at https://doi.org/10.1016/j.annonc.2023.08.006). To assess the potential influence of treatment received on patient survival in the three molecular subgroups, an interaction test between mutational status (DWT, BRAFV600E, and KRAS) and treatment was performed and found to be nonsignificant (TTR Pinteraction = 0.08; OS Pinteraction = 0.42), leading to the conclusion that the impact of KRAS and BRAFV600E mutations on TTR and OS could be analyzed independently of the treatment received.
Outcome predictors in patients with MSS tumors
As in the whole population, when considering patients with MSS tumors, TTR was significantly shorter in patients with KRAS MT tumors [hazard ratio (HR) = 1.31, 95% CI 1.20-1.42; P < 0.0001] and in patients with BRAFV600E MT tumors (HR = 1.58, 95% CI 1.36-1.83; P < 0.0001) compared with patients with DWT tumors (Figure 2A). The percentage of patients who were alive and without disease recurrence at 5 years was 72.9%, 66.3%, and 61.8% in the DWT, KRAS MT, and BRAF MT groups, respectively.
Figure 2. The Kaplan—Meier curve for time to recurrence (TTR), overall survival (OS), and survival after recurrence (SAR) for (A, C, and E) the microsatellite-stable (MSS) population and (B, D, and F) the microsatellite unstable (MSI-high) population by mutational status.
CI, confidence interval; dMMR, deficient mismatch repair; HR, hazard ratio; MT, mutated; WT, wild type.
OS was significantly shorter in patients with KRAS MT tumors (HR = 1.35, 95% CI 1.23-1.49; P < 0.0001) and in patients with BRAF MT tumors (HR = 2.06, 95% CI 1.78-2.39; P < 0.0001; Figure 2B). The rate of patients alive at 8 years was 73.6%, 67.2%, and 58.9% in the DWT, KRAS MT, and BRAF MT groups, respectively.
SAR was also significantly shorter in patients with KRAS and BRAF MT tumors. The median SAR was 2.7, 2.3, and 0.9 year(s) in the DWT, KRAS MT (HR = 1.25, 95% CI 1.12-1.39; P < 0.0001), and BRAF MT (HR = 2.87, 95% CI 2.44-3.38; P < 0.0001) groups, respectively (Figure 2C).
In a multivariable analysis, all tested variables were significantly associated with shorter OS and all tested variables, except age and primary tumor location, with shorter TTR (Supplementary Table S5, available at https://doi.org/10.1016/j.annonc.2023.08.006). Considering SAR, characteristics including BRAF MT (P < 0.0001), KRAS MT (P < 0.02), high-grade tumors (P < 0.001), pT4 (P < 0.01), and pN2 (P < 0.0001) remained associated with shorter SAR.
Outcome predictors for patients with MSI-H tumors
In the MSI-H cohort, there was no difference for TTR in patients with KRAS MT (HR = 0.99; P = 0.97) or BRAF MT tumors (HR = 0.98; P = 0.91) as compared with DWT patients (Figure 2D).
The percentage of patients who were alive and without disease recurrence at 5 years was 75.4%, 75.8%, and 76.1% in the DWT, KRAS MT, and BRAF MT groups, respectively.
Considering OS, as shown in Figure 2E, no significant impact on patient outcome was seen for those with KRAS MT tumors (HR = 1.18; P = 0.39), but OS was shorter in the cohort with BRAF MT as compared with patients with DWT tumors (HR = 1.36; P = 0.042). The rate of patients alive at 8 years was 76.9%, 74.6%, and 71.6% in the DWT, KRAS MT, and BRAF MT groups, respectively.
The SAR was shorter in patients with KRAS and BRAF MT tumors. The median SAR was 1.9, 1.3, and 0.7 years in DWT, KRAS MT (HR = 1.52, 95% CI 0.97-2.37; P = 0.07), and BRAF MT (HR = 1.99, 95% CI 1.38-2.89; P < 0.001) populations, respectively (Figure 2F).
In a multivariable analysis among patients with MSI-H tumors, PS, tumor grade, pT stage, and pN stage were associated with significantly shorter TTR. Similar results were obtained for the association of OS with pT4 stage (P = 0.0003) and pN2 stage (P < 0.0001), as well as with age >70 years (P = 0.004), female sex (P = 0.01), and Eastern Cooperative Oncology Group (ECOG) PS 1 or 2 (P = 0.01), as these factors were all associated with shorter OS. No association between KRAS and BRAFV600E mutational status and TTR or OS was observed in the multivariable analysis for patients with MSI-H. However, SAR was worse in both subgroups of patients with KRAS MT (HR = 1.81; 95% CI 1.11-2.93, P = 0.017) and BRAF MT (HR = 1.99; 95% CI 1.30-3.03, P < 0.01) tumors as compared with the cohort of patients with DWT tumors (Supplementary Table S6, available at https://doi.org/10.1016/j.annonc.2023.08.006).
KRAS exon 2 submutations (G12C, G12D, G12V, G13D, and others) and prognosis
Among the 2889 KRAS MT MSS tumors, 9.3% were p.G12C, 32.4% were p.G12D, 21.3% were p.G12V, 16.2% were other mutations, and 20.8% were p.G13D as shown in Figure 1.
Among the 175 KRAS MT MSI-H tumors, 2.3% were p.G12C, 36.0% were p.G12D, 6.9% were p.G12V, 12.0% were other mutations, and 42.9% were p.G13D (Figure 1).
The association of KRAS submutations and TTR, OS, and SAR was only assessed in patients with MSS tumors and not in those with MSI-H tumors due to the limited number of patients in this subgroup.
In patients with MSS tumors, all codon 12 alterations, p.G13D mutations, and BRAFV600E mutations were associated with shorter TTR (Figure 3A) and OS (Figure 3B), except for the cohort with p.G12V where the differences in OS were nonstatistically significant (P = 0.057).
Figure 3. Forest plot for (A) time to recurrence (TTR), (B) overall survival (OS), and (C) survival after recurrence (SAR) for the microsatellite-stable (MSS) population depending on mutations. Kaplan Meier curves for (D) time to recurrence (TTR), (E) overall survival (OS), and (F) SUrvival after recurrence (SAR) for the micro-satellite stable (MSS) population depending on mutations.
HRs and 95% CIs were calculated using the adjusted Cox proportional hazard model. Solid circles represent HR, whereas open-ended horizontal lines represent the 95% CIs.
CI, confidence interval; HR, hazard ratio; MT, mutated; WT, wild type.
When considering SAR alone, p.G12C, p.G12D, p.G13D, and BRAF V600E mutations were associated with shorter SAR. Figure 3C summarizes the impact of individual mutations on SAR.
Figure 3D-F show the Kaplan—Meier curves of TTR, OS, and SAR for each individual mutation with 5-year TTR, 8-year OS, and 2-year SAR percentages reported on the figures.
DISCUSSION
To our knowledge, our report is the largest series evaluating the potential prognostic value of KRAS exon 2 submutations and BRAFV600E stratified by MSI/MSS status in patients with stage III CC. With >7000 patients analyzed, our results show that both KRAS exon 2 and BRAFV600E mutations are important prognostic factors for both TTR and OS in stage III MSS CC but not in their MSI counterpart. Indeed, 5-year disease recurrence rates were 27.1% in DWT, 33.7% in KRAS MT, and 38.2% in BRAF MT patients with MSS tumors but these rates were similar in all molecular subgroups in patients with MSI-H (24.6%, 24.2%, and 23.9%, respectively). Patients with MSI-H will be treated differently in the future as a result of the efficacy of ICI for this specific subgroup, but our results underline the unmet need for more efficient treatments for patients with KRAS MT and BRAF MT MSS exhibiting particularly high rates of disease recurrence.
In this large cohort of stage III patients, we found a higher frequency of BRAF MT (40.6% versus 7.7%) and fewer KRAS MT (18% versus 36%) in MSI-H CCs as compared with MSS CCs. When looking at KRAS exon 2 submutations, the druggable G12C and G12D mutations were found in 9.3% and 32.4% of patients with MSS tumors, respectively, corresponding to prior estimates (ref).3 Although KRAS codon 13 (p.G13D) mutations were found in a similar percentage of MSI-H and MSS tumors (~8%), codon 12 mutations were very rare in patients with MSI-H tumors, especially when considering p.G12C (2.3%) and p.G12V (6.9%).
Considering clinicopathological characteristics, we found that older age, female sex, PS > 0, N2 stage, poor differentiation, and proximal location were more common in patients with MSS BRAF MT tumors as compared with DWT tumors. Similar characteristics were observed for BRAF MT tumors in patients with MSI-H CC except for those with N2 stage, as compared with the DWT cohort. No major differences in baseline characteristics were observed for patients with KRAS MT tumors as compared with DWT in both MSS and MSI-H patient groups, except for more frequent proximal location in KRAS-mutant MSS tumors (Table 1).
We found increased recurrence rates in both patients with KRAS MT and patients with BRAF MT with MSS CC. Large studies of patients with nonmetastatic CRC have given conflicting results regarding the prognostic value of KRAS and BRAFV600E mutations on disease recurrence.12-22 In addition to differences in the conduct of the analyses (mutant versus wild type), this may be due to heterogeneity in the study’s populations, with respect to the frequencies of factors such as tumor site (rectal and colon), tumor stage (I to III), and adjuvant treatment received (surgery alone, fluoropyrimidines alone, and fluoropyrimidines + oxaliplatin). When looking at TTR in our homogeneous cohort of patients with stage III CC, we found a clinically relevant association of both KRAS and BRAF mutations with disease recurrence. These results were confirmed after adjustment for key prognostic factors for disease recurrence in stage III CC such as age, sex, ECOG PS, pT stage, pN stage, primary tumor location, and tumor grade. Similar results were found for OS in the MSS group, with 8-year OS rates being inferior in patients with KRAS and BRAF MT tumors, respectively, as compared with DWT tumors. Finally, similar results were found when the analyses were restricted to patients only treated with a standard fluoropyrimidine and oxaliplatin combination (Supplementary Figure S2, available at https://doi.org/10.1016/j.annonc.2023.08.006).
Although a slightly higher rate of recurrence was observed for patients with BRAF MT tumors in the first 2 years after surgery, no long-term difference in disease recurrence in KRAS MT, BRAF MT, and DWT patient cohorts was found in patients in the MSI-H group. The very specific immune microenvironment seen in these MSI-H tumors and their relatively good prognosis may be responsible for the lack of prognostic value of KRAS exon 2 and BRAFV600E mutations in these patients. When considering patient OS in the MSI-H group, a trend toward poorer OS in KRAS MT and BRAF MT tumors was seen, but this did not reach significance in the multivariable analysis.
Previous studies have suggested that not all KRAS exon 2 submutations were associated with poorer TTR, suggesting differences between individual KRAS submutations. In some studies, only the p.G12V (RASCAL) mutation was associated with poorer outcome and in others, KRAS mutations at codon 12, but not at codon 13, negatively affected patients’ outcome.15,18,22,23 When looking at KRAS exon 2 submutations and their association with TTR in the present work, we found similar outcomes for all KRAS exon 2 submutations in patients with MSS tumors whereby all were associated with a poorer TTR with HRs ranging from 1.23 to 1.37. We also found that all KRAS exon 2 submutations were associated with a poorer OS with HRs ranging from 1.18 to 1.44 for individual mutations. However, statistical significance was not reached for the pG12V submutation (P = 0.058).
Finally, we also analyzed SAR, which refers to the survival time between disease recurrence and death. Here we found, as expected, worse survival for cohorts with KRAS MT and BRAF MT tumors as compared with DWT tumors in the MSS group, but similar results were also found in patients with MSI-H as reported previously for BRAF MT.24 This finding suggests that KRAS and BRAFV600E mutations are poor prognostic markers after disease recurrence regardless of MSI status. When looking at KRAS exon 2 submutations and SAR, both KRAS p.G12C and p.G13D cohorts seem to exhibit worse survival as compared with other KRAS exon 2 submutations or DWT cohorts. This has been recently reported in other series outside clinical trials.25 It is important to point out here that due to the timing of the adjuvant studies included in the pooled analyses, patients did not receive any specific treatment linked to their molecular profile (i.e. BRAF inhibitors, KRAS G12C inhibitors, or immune checkpoint inhibitors) for their advanced disease.
Important strengths of our study are the large size of our clinical cohort that pooled seven recent prospective randomized controlled trials with adjuvant treatments, prospective tissue collection, long follow-up, and high-quality clinical annotations. However, this work also has several limitations including the retrospective design of these pooled analyses, the absence of expanded RAS mutation profiling as recommended by the European Society for Medical Oncology and National Comprehensive Cancer Network guidelines in the metastatic setting, and the population coming from clinical trials that may not completely reflect real-world patient populations.
In conclusion, this study represents the largest analysis to assess the impact of KRAS exon 2 and BRAFV600E mutations on oncological outcomes in patients with surgically resected CC. We demonstrated the prognostic value of these mutated oncogenes on TTR, OS, and SAR in patients with MSS, and on SAR alone for patients with MSI-H. In patients with MSS tumors, all KRAS exon 2 submutations were associated with similar prognosis in nonmetastatic patients, yet both KRAS p.G12C and p.G13D are associated with poor survival after disease recurrence. Our results suggest that KRAS and BRAFV600E mutations have to be tested and included as stratification variables in future adjuvant clinical trials dedicated to patients with stage III CC. In our opinion, immediate testing of the surgical specimen for KRAS and BRAF mutational status should also be considered in routine practice, as this can add prognostic information for each individual patient with resected stage III CC and quickly inform an appropriate first-line therapy in case of disease recurrence. In addition to perioperative immunotherapy, which has shown impressive results in patients with nonmetastatic MSI-H26,27 and is currently tested in phase III trials, our results support neoadjuvant and adjuvant trials testing for specific RAS- and BRAF-targeted agents, in an effort to overcome the poor prognosis associated with these mutations in patients with MSS. Some trials testing encorafenib and cetuximab in the neoadjuvant setting are currently ongoing for BRAFV600E-mutated tumors (EUDRACT number 2021-000828-35 and NCT05510895) and others testing panitumumab in KRAS G12C-mutated tumors are in development.
Supplementary Material
ACKNOWLEDGEMENTS
The authors thank ACCENT and IDEA consortia for their participation in the study and all the cooperative groups, patients, and family involved in the seven trials analyzed in this work.
Footnotes
DISCLOSURE
JT has received honoraria for speaker or advisory role from Astellas, Roche, Merck Serono, Amgen, Servier, Pierre Fabre, BMS, Astra Zeneca, Novartis, Takeda, and MSD. FAS reports consulting/advisory funds from Guardant Health and Roche/Genentech; and patent-related royalties from Roche Diagnostics. QS reports consulting/advisory role from Yiviva Inc, Boehringer Ingelheim Pharmaceuticals, Inc, Regeneron Pharmaceuticals, Inc., Hoosier Cancer Research Network, Kronos Bio, and Mirati Therapeutics Inc; honorarium/speaker role from Chugai Pharmaceutical Co., Ltd (to self); and research funds from Celgene/BMS, Roche/Genentech, Janssen, and Novartis (to institution). RMG reports the receipt of consulting fees from AbbVie, Adaptimmune, Astra Zeneca, Bayer, Compass Therapeutics, Focal Medical, Haystack Oncology, Innovative Cellular Therapeutics, Inspirna, IQVIA, Merck, Sorrento, Taiho, and Takeda; also reports stock options from Compass Therapeutics, Haystack, and Focal Medical; and has served as a paid expert witness for Taiho and Genentech. TJG reports consulting/advisory role from Pfizer/Array, Tempus Labs, and BillionToOne. TA reports honoraria from Amgen, Aptitude Health, AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Merck Sharp & Dohme, Pierre Fabre, Roche/Vantana, Sanofi, and Servier; consulting or advisory roles for Astellas Pharma, Bristol-Myers Squibb, GamaMabs Pharma, Gilead, GlaxoSmithKline, Gritstone Oncology, Merck Sharp & Dohme, Seagen, and Servier; participating in Transgene Speaker’s Bureaus for Bristol-Myers Squibb, Merck Sharp & Dohme, Seagen, and Servier; research funding from Bristol-Myers Squibb and Merck Sharp & Dohme; travel and accommodation expenses covered by Merck Sharp & Dohme and Bristol-Myers Squibb; and nonremunerated activities for the ARCAD Foundation and GERCOR Group. SRA serves as a member of several data monitoring committees for Merck. LS reports consulting role for Genor Biopharma, Ltd. PLP has received honoraria for speaker or advisory role from Sanofi, Merck Serono, Amgen, Servier, Pierre Fabre, and Biocartis; is also the founder of METHYS Dx. SL reports research funding (to institution) from Amgen, Merck Serono, Bayer, Roche, Eli Lilly, AstraZeneca, and Bristol Myers Squibb; personal honoraria as an invited speaker from Roche, Eli Lilly, Bristol Myers Squibb, Servier, Merck Serono, Pierre Fabre, GlaxoSmithKline, and Amgen; participation in advisory board for Amgen, Merck Serono, Eli Lilly, AstraZeneca, Incyte, Daiichi-Sankyo, Bristol Myers Squibb, Servier, and Merck Sharp & Dohme. TY reports honoraria from Bayer, Chugai, Merck Biopharma, MSD, Ono, and Takeda; consulting role from Sumitomo Corp.; research grants from Taiho, Ono, Chugai, Amgen, MSD, Daiichi-Sankyo, Eisai, FALCO Biosystems, Genomedia, Molecular Health, Nippon Boehringer Ingelheim, Pfizer, Roche Diagnostics, Sysmex, and Sanofi. All other authors have declared no conflicts of interest.
REFERENCES
- 1.Van Cutsem E, Cervantes A, Adam R, et al. ESMO consensus guidelines for the management of patients with metastatic colorectal cancer. Ann Oncol. 2016;27(8):1386–1422. [DOI] [PubMed] [Google Scholar]
- 2.Cervantes A, Adam R, Roselló S, et al. Metastatic colorectal cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol. 2023;34(1):10–32. [DOI] [PubMed] [Google Scholar]
- 3.Taieb J, Jung A, Sartore-Bianchi A, et al. The evolving biomarker landscape for treatment selection in metastatic colorectal cancer. Drugs. 2019;79(13):1375–1394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Benson AB, Venook AP, Al-Hawary MM, et al. Colon cancer, Version 2. 2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Cancer Netw. 2021;19(3):329–359. [DOI] [PubMed] [Google Scholar]
- 5.Kopetz S, Grothey A, Yaeger R, et al. Encorafenib, binimetinib, and cetuximab in BRAF V600E-mutated colorectal cancer. N Engl J Med. 2019;381(17):1632–1643. [DOI] [PubMed] [Google Scholar]
- 6.Andre T, Shiu K-K, Kim TW, et al. Pembrolizumab in microsatellite-instability-high advanced colorectal cancer. N Engl J Med. 2020;383(23):2207–2218. [DOI] [PubMed] [Google Scholar]
- 7.Ciardiello D, Maiorano BA, Martinelli E. Targeting KRASG12C in colorectal cancer: the beginning of a new era. ESMO Open. 2023;8(1):100745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ahn D, Walden D, Bekaii-Saab T. Tucatinib: an investigational novel therapeutic agent for the treatment of HER-2 colorectal cancer. Expert Opin Investig Drugs. 2022;31(5):437–441. [DOI] [PubMed] [Google Scholar]
- 9.Strickler JH, Yoshino T, Graham RP, Siena S, Bekaii-Saab T. Diagnosis and treatment of ERBB2-positive metastatic colorectal cancer: a review. JAMA Oncol. 2022;8(5):760–769. [DOI] [PubMed] [Google Scholar]
- 10.Hong DS, DuBois SG, Kummar S, et al. Larotrectinib in patients with TRK fusion-positive solid tumours: a pooled analysis of three phase 1/2 clinical trials. Lancet Oncol. 2020;21(4):531–540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Argilés G, Tabernero J, Labianca R, et al. Localised colon cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2020;31(10):1291–1305. [DOI] [PubMed] [Google Scholar]
- 12.Andreyev HJ, Norman AR, Cunningham D, et al. Kirsten Ras mutations in patients with colorectal cancer: the “RASCAL II” study. Br J Cancer. 2001;85(5):692–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fariña-Sarasqueta A, van Lijnschoten G, Moerland E, et al. The BRAF V600E mutation is an independent prognostic factor for survival in stage II and stage III colon cancer patients. Ann Oncol. 2010;21(12):2396–2402. [DOI] [PubMed] [Google Scholar]
- 14.Hutchins G, Southward K, Handley K, et al. Value of mismatch repair, KRAS, and BRAF mutations in predicting recurrence and benefits from chemotherapy in colorectal cancer. J Clin Oncol. 2011;29(10):1261–1270. [DOI] [PubMed] [Google Scholar]
- 15.Gavin PG, Colangelo LH, Fumagalli D, et al. Mutation profiling and microsatellite instability in stage II and III colon cancer: an assessment of their prognostic and oxaliplatin predictive value. Clin Cancer Res. 2012;18(23):6531–6541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Imamura Y, Morikawa T, Liao X, et al. Specific mutations in KRAS codons 12 and 13, and patient prognosis in 1075 BRAF wild-type colorectal cancers. Clin Cancer Res. 2012;18(17):4753–4763. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Morikawa T, Kuchiba A, Qian ZR, et al. Prognostic significance and molecular associations of tumor growth pattern in colorectal cancer. Ann Surg Oncol. 2012;19(6):1944–1953. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Blons H, Emile JF, Le Malicot K, et al. Prognostic value of KRAS mutations in stage III colon cancer: post hoc analysis of the PETACC8 phase III trial dataset. Ann Oncol. 2014;25(12):2378–2385. [DOI] [PubMed] [Google Scholar]
- 19.André T, de Gramont A, Vernerey D, 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. 2015;33(35):4176–4187. [DOI] [PubMed] [Google Scholar]
- 20.Taieb J, Le Malicot K, Shi Q, et al. Prognostic value of BRAF and KRAS mutations in MSI and MSS stage III colon cancer. J Natl Cancer Inst. 2017;109(5):djw272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sinicrope FA, Mahoney MR, Yoon HH, et al. Analysis of molecular markers by anatomic tumor site in stage III colon carcinomas from adjuvant chemotherapy trial NCCTG N0147 (Alliance). Clin Cancer Res. 2015;21(23):5294–5304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sinicrope FA, Shi Q, Smyrk TC, et al. Molecular markers identify subtypes of stage III colon cancer associated with patient outcomes. Gastroenterology. 2015;148(1):88–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ogino S, Shima K, Meyerhardt JA, et al. Predictive and prognostic roles of BRAF mutation in stage III colon cancer: results from intergroup trial CALGB 89803. Clin Cancer Res. 2012;18(3):890–900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Taieb J, Shi Q, Pederson L, et al. Prognosis of microsatellite instability and/or mismatch repair deficiency stage III colon cancer patients after disease recurrence following adjuvant treatment: results of an ACCENT pooled analysis of seven studies. Ann Oncol. 2019;30(9):1466–1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fakih M, Tu H, Hsu H, et al. Real-world study of characteristics and treatment outcomes among patients with KRAS p.G12C-mutated or other KRAS mutated metastatic colorectal cancer. Oncologist. 2022;27(8):663–674. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ludford K, Ho WJ, Thomas JV, et al. Neoadjuvant pembrolizumab in localized microsatellite instability high/deficient mismatch repair solid tumors. J Clin Oncol. 2023;41(12):2181–2190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Taieb J, Svrcek M, Cohen R, Basile D, Tougeron D, Phelip J-M. Deficient mismatch repair/microsatellite unstable colorectal cancer: diagnosis, prognosis and treatment. Eur J Cancer. 2022;175:136–157. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.