Abstract
Background
HER2 amplification is demonstrated to influence response to therapy and outcomes in metastatic colorectal cancer (mCRC), but its role in resectable CRC remains unclear. This study investigated the prognostic impact of WES-defined ERBB2 amplification and its association with adjuvant chemotherapy (ACT) efficacy and circulating tumor DNA (ctDNA) in resectable CRC.
Methods
Data from 3607 patients with resectable CRC from the GALAXY study were analyzed. HER2 amplification was defined using ERBB2 copy number alterations based on whole exome sequencing of tumor specimens. We analyzed disease-free survival (DFS) in WES-defined ERBB2 amplified CRC, evaluating ACT benefit and the correlation between ERBB2 copy number and ctDNA status.
Results
No significant difference in DFS was observed between those with and without WES-defined ERBB2 amplification in a univariate (P = 0.0659) or a multivariate analysis (P = 0.2300). Among patients with pathological stage III, ACT was associated with improved DFS in the non-amplified subgroup (HR 0.67, P = 0.0002), particularly in the subgroup of patients with early post-operative (2–10 weeks following surgical resection) ctDNA-positivity (HR 0.27, P < 0.0001), whereas no clear association was detected in the small amplified subgroup; however, the ACT-by-ERBB2 amplification interaction was not statistically significant. While WES-defined ERBB2 amplification correlated with higher early post-operative ctDNA positivity (P = 0.0486), ERBB2 copy number did not associate with recurrence or levels. However, early post-operative ctDNA positivity strongly predicted poor DFS in patients with WES-defined ERBB2 amplification (HR 15.26, P < 0.0001).
Conclusions
WES-defined ERBB2 amplification was not identified as a significant adverse prognostic factor in resectable CRC, unlike early post-operative ctDNA status. The limited number of amplified cases precluded a reliable assessment of whether WES-defined ERBB2 amplification modifies the association between ACT and DFS.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1007/s10147-026-03161-5.
Keywords: HER2 amplification, Resectable colorectal cancer, Prognostic marker, Relevance to adjuvant chemotherapy efficacy, ctDNA
Introduction
Amplification and/or overexpression of human epidermal growth factor receptor 2 (HER2) gene, referred to as HER2 positivity, is relatively uncommon in colorectal cancer (CRC), with an estimated prevalence of approximately 2–4% [1, 2]. Data from population-based and trial cohorts report HER2 positivity in approximately 2–5% of metastatic CRC (mCRC), compared with ~ 1–2% of resected stage II–III tumors [3–5]. In mCRC, numerous studies have demonstrated that HER2 positivity influences response to therapy and clinical outcomes. In particular, patients with HER2-positive, wild-type RAS mCRC show poor response to anti-EGFR therapy and shorter progression-free survival (PFS) [6, 7]. HER2 positivity has also been associated with unfavorable overall survival (OS) in mCRC [7]. Recently, HER2-targeted therapy has emerged as a promising treatment strategy for mCRC. The efficacy of dual HER2 blockade with trastuzumab-based regimens has been confirmed in multiple clinical trials [8–12]. In Japan, pertuzumab plus trastuzumab was approved in 2022 for HER2-positive mCRC, and in 2023, the combination of tucatinib and trastuzumab received regulatory approval. These advances have firmly established HER2 as an important therapeutic target in mCRC.
In contrast, the clinical significance of HER2 amplification/overexpression in resectable CRC remains unclear. Although several studies have suggested its associations with postoperative recurrence, prognosis, and the efficacy of adjuvant chemotherapy (ACT) [13–15], evidence is still insufficient to determine whether HER2 positivity serves as an independent prognostic or predictive factor in this setting. This uncertainty likely reflects both its low prevalence and the fact that HER2 testing is not routinely performed in stage I–III CRC in clinical practice.
With recent advances in next-generation sequencing (NGS), genetic alterations and copy number aberrations—including KRAS mutations and HER2 amplification—can now be detected within a single assay. Furthermore, assessment of HER2 amplification in CRC using NGS-based assays has shown high concordance with immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) results [16], supporting its reliability for detecting HER2 amplification. These advances provide an opportunity to systematically evaluate the clinical relevance of WES-defined ERBB2 amplification in large cohorts of patients with resectable CRC.
The GALAXY study, which is the observational arm of the ongoing CIRCULATE-Japan study (UMIN000039205), enrolled more than 6000 patients with resectable, clinical stage (cStage) II-IV CRC. Whole-exome sequencing (WES) of tumor tissues was performed for each patient to design personalized, tumor-informed circulating tumor DNA (ctDNA) assays for evaluating recurrence risk. In this study, we leveraged this large and homogeneous data from the GALAXY study and evaluated the clinical significance of WES-defined ERBB2 amplification in resectable CRC.
Methods
Patients and study design
This study is a secondary analysis using data from the GALAXY study, the observational arm of the prospective, multicenter CIRCULATE-Japan study. The GALAXY study is a nationwide registry that prospectively monitors ctDNA status after complete surgical resection in patients with cStage II-IV CRC [17, 18]. It serves to screen patients for ctDNA-based molecular residual disease (MRD) status and subsequently assign patients to one of the two randomized ctDNA-guided interventional phase 3 trials: ALTAIR (treatment escalation in MRD-positive patients) and VEGA (treatment de-escalation in MRD-negative patients) [17, 19]. The detailed study protocol has been previously reported [19]. In addition, the eligibility and exclusion criteria, procedures for sample collection and testing, administration of ACT, definition of DFS, and methods for recurrence assessment have been described in the interim analyses [17, 18]. A total of 6,061 patients were enrolled between May 8, 2020, and March 31, 2024.
For this current study, we excluded those who withdrew consent, received neoadjuvant therapy, had an uncertain date of surgery, or enrolled in the interventional ALTAIR or VEGA studies (Fig. 1). This resulted in a cohort of 3,607 patients with cStage II-III and pathologic stage (pStage) I-III CRC for this analysis. In each subgroup analysis, patients with missing data for the analyzed variables were excluded. No patients in this analysis received anti-EGFR or anti-HER2 therapies.
Fig. 1.
CONSORT diagram illustrating the inclusion of patients
Personalized, tumor-informed ctDNA analysis
Early post-operative ctDNA-based MRD status during 2–10 weeks following surgical resection was assessed as part of the GALAXY study using a clinically validated, personalized, tumor-informed 16-plex mPCR-NGS assay (Signatera™, Natera, Inc.), as previously described [20–22]. Briefly, a set of up to 16 single nucleotide variants (SNVs) was identified from paired WES data generated from fixed-formalin paraffin-embedded (FFPE) tumor tissue and normal blood samples. These SNVs were tracked in the associated patient’s plasma using a multiplex PCR-based NGS approach. Samples with at least 2 patient-specific SNVs detected above a predefined threshold were considered ctDNA-positive (MRD-positive). ctDNA concentrations (levels) were reported as mean tumor molecules per mL of plasma (MTM/mL).
Definition of WES-based ERBB2 amplification
ERBB2 copy number was estimated from WES-based log2 fold-change values using the following formula: CN = 2 × 2(log2FC). The cut-off for ERBB2 amplification as defined by copy number varies between 3 and 8 [23]. In the WES analysis of our cohort, the lowest ERBB2 copy number among tumors showing copy-number gain was 4.59. Based on this WES-defined ERBB2 amplification was defined as a ERBB2 copy number greater than 4.0, based on WES analysis of tumor specimens. We were unable to evaluate tumor purity, ploidy, or the focality of amplification in the current dataset.
Statistical analysis
Categorical variables were compared using the chi-square or Fisher’s exact test, and continuous variables using the Mann–Whitney U test. Survival analyses were conducted in R version 4.4.3 with the survival (version 3.8.3) and survminer (version 0.5.0) packages. Survival curves were estimated by the Kaplan–Meier method, group differences evaluated with the log-rank test, and hazard ratios calculated using Cox proportional hazards models and Firth corrected-Cox from package coxphf (version 1.13.4) was used when event count is low. Multivariate Cox models were used to identify prognostic factors associated with DFS. Box plots were generated with ggplot2 (version 3.5.2) and ggpubr (version 0.6.1), and data manipulation performed with dplyr (version 1.1.4). P values < 0.05 were considered statistically significant.
A landmark analysis, which offset the disease-free survival (DFS) starting point by 70 days, was performed using Kaplan-Meier survival estimates and a Multivariate Cox proportional hazards model incorporating early post-operative MRD status. This 70 days offset was implemented to prevent immortal time bias (guarantee-time bias). By applying this landmark, we ensured that patients who experienced rapid recurrence or death prior to the completion of the standard post-operative MRD testing window were appropriately excluded, allowing for an unbiased comparison between the MRD-positive and MRD-negative cohorts. The number of patients and DFS events excluded before the landmark is reported in Supplementary Table 1.
The sensitivity analysis was performed for association of WES-defined ERBB2 amplification with DFS in univariate and multivariate analysis, and efficacy of ACT based on WES-defined ERBB2 amplification status using the ERBB2 copy number between 3 and 20.
Results
Patient characteristics
The baseline patient and tumor characteristics are shown in Table 1. Among 3,607 patients, 59 (1.6%) had WES-defined ERBB2 amplified tumors, i.e. ERBB2 copy number greater than 4.0. We noted that the cutoff at ERBB2 copy number 3 or 4 would give the same results in this cohort (Fig. S1). The WES-defined ERBB2 amplified group included a higher proportion of pStage III cases (61.0% vs. 47.1%; P = 0.0358) and lower frequencies of RAS mutations (8.5% vs. 38.4%; P < 0.0001), BRAF V600E mutation (0% vs. 8.4%; P = 0.0083), and MSI-high status (0% vs. 11.1%; P = 0.0014). Interestingly, early post-operative MRD positivity during 2–10 weeks post-surgery was higher in the WES-defined ERBB2 amplified group (18.6% vs. 9.9%; P = 0.0486).
Table 1.
Baseline demographic and clinical characteristics of patients with and without WES-defined ERBB2 amplification
| Characteristic | WES-defined ERBB2 amp (−) (N = 3548) 98.4% |
WES-defined ERBB2 amp (+) (N = 59) 1.6% |
P value (Fisher or Wilcoxon) |
|
|---|---|---|---|---|
| Sex | Male | 1844 (52.0%) | 33 (55.9%) | 0.6002 |
| Female | 1704 (48.0%) | 26 (44.1%) | ||
| Age, years | median (IQR) | 69 (59–75) | 66 (58–73) | 0.3816 |
| Primary tumor location | Right (C-T) | 1260 (35.5%) | 14 (23.7%) | 0.1170 |
| Left (D-RS) | 1724 (48.6%) | 32 (54.2%) | ||
| Rectum (Ra, Rb) | 564 (15.9%) | 13 (22.0%) | ||
| Histological type | Differentiated | 3362 (94.8%) | 58 (98.3%) | 0.2628 |
| Poorly differentiated/ Signet-ring cell | 115 (3.2%) | 0 (0%) | ||
| Other/Unknown | 71 (2.0%) | 1 (1.7%) | ||
| Pathologic stage | Stage I | 675 (19.0%) | 7 (12%) | 0.0358 |
| Stage II | 1202 (33.9%) | 16 (27%) | ||
| Stage III | 1671 (47.1%) | 36 (61%) | ||
| RAS mutations | Wild type | 1919 (54.1%) | 52 (88.1%) | < 0.0001 |
| Mutant | 1364 (38.4%) | 5 (8.5%) | ||
| Unknown | 265 (7.5%) | 2 (3.4%) | ||
| BRAF V600E mutation | No | 2982 (84.0%) | 57 (96.6%) | 0.0083 |
| Yes | 297 (8.4%) | 0 (0%) | ||
| Unknown | 269 (7.6%) | 2 (3.4%) | ||
| MSI | MSI-high | 393 (11.1%) | 0 (0%) | 0.0014 |
| MSS | 2828 (79.7%) | 57 (96.6%) | ||
| Unknown | 327 (9.2%) | 2 (3.4%) | ||
| Early post-operative MRD status (2–10 weeks post-surgery) | Negative | 2782 (78.4%) | 43 (72.9%) | 0.0486 |
| Positive | 352 (9.9%) | 11 (18.6%) | ||
| Unknown | 414 (11.7%) | 5 (8.5%) |
IQR Interquartile range; Right (C-T), cecum to transverse colon (proximal colon); Left (D-RS), descending colon to Rectosigmoid (distal colon); Ra, upper rectum; Rb, lower rectum; MSI, microsatellite instability; MSS, microsatellite stable; MRD, molecular residual disease. P value was calculated using Fisher’s exact test among non-missing categories
Association of WES-defined ERBB2 amplification status with DFS
Among 3152 patients with available DFS data, the median follow-up was 783 days (IQR: 372–1107) as of March 12, 2025. Recurrence or death occurred in 12 of 48 patients with WES-defined ERBB2 amplification (25.0%) and in 520 of 3,104 patients without amplification (16.8%). Comparison of recurrence patterns between the WES-defined ERBB2 amplified and non-amplified groups showed no remarkable differences in the sites of recurrence (Supplementary Table 2). The 3 year DFS was 71.7% in patients with WES-defined ERBB2 amplification and 81.1% in patients without amplification, but the difference was not statistically significant (hazard ratio [HR]: 1.70; 95% confidence interval [CI]: 0.96–3.01; P = 0.0659) (Fig. 2a). Similarly, analyses stratified by pStage showed no significant association between WES-defined ERBB2 amplification and DFS (Fig. 2b–d).
Fig. 2.
Association of WES-defined ERBB2 amplification with DFS. Kaplan–Meier estimates of DFS stratified by WES-defined ERBB2 amplification status among all patients with available DFS data (N = 3152) (a), patients with pStage I (N = 590) (b), pStage II (N = 1037) (c), or pStage III (N = 1525) (d) disease. DFS, disease-free survival; pStage, pathological stage
To further assess the prognostic significance of WES-defined ERBB2 amplification while accounting for other clinicopathological factors, a multivariable analysis was conducted in 2487 patients with complete data and who were event-free until the 70 days landmark. In the multivariate analysis, WES-defined ERBB2 amplification was not observed to be a significant adverse prognostic factor (HR: 1.51; 95% CI: 0.77–2.96; P = 0.2300; Fig. 3), whereas early post-operative MRD positivity was the strongest adverse factor of DFS (HR: 7.33; 95% CI: 5.96–9.03; P < 0.0001; Fig. 3). Other statistically significant adverse prognostic factors included advanced pStage (II vs. I: HR: 2.38; 95% CI: 1.37 − 4.11; P = 0.0019; III vs. I: HR: 4.34, 95% CI: 2.56 − 7.37, P < 0.0001), MSS (HR: 3.95; 95% CI: 2.14–7.28; P < 0.0001), BRAF V600E mutation (HR: 2.53; 95% CI: 1.59–4.03; P < 0.0001), RAS mutations (HR: 1.52; 95% CI: 1.24–1.87; P < 0.0001), male sex (HR: 1.32; 95% CI: 1.08–1.62; P = 0.0059), and rectal location (HR: 1.33; 95% CI: 1.04–1.69; P = 0.0233) (Fig. 3). Considering that adjustment for early post-operative MRD status may affect the interpretation of the prognostic value of WES-defined ERBB2 amplification, multivariable analysis excluding MRD status was also conducted; however, WES-defined ERBB2 amplification was not significantly associated with DFS in the model. (Supplementary Fig. 2).
Fig. 3.
Multivariable analysis of DFS among patients with complete clinicopathologic data available (N = 2487). Landmark analysis was performed
Efficacy of ACT stratified by WES-defined ERBB2 amplification and MRD status
Because ACT is recommended as the standard of care for pStage III CRC, we analyzed its benefit in that patient subgroup upon stratifying by WES-defined ERBB2 amplification status. The baseline characteristics according to ACT receipt among pStage III patients are shown in Table 2 for the overall cohort and in Supplementary Tables 3 and 4 for cohorts stratified by WES-defined ERBB2 amplification status. Age and early post-operative MRD status within 2–10 weeks post-surgery differed significantly between the ACT-treated and untreated groups in the overall cohort and in the non-amplified subgroup. However, no statistically significant differences in baseline characteristics were observed in the WES-defined ERBB2 amplified subgroup. The distribution of ACT regimens and treatment durations are shown in Supplementary Tables 5 and 6.
Table 2.
Baseline demographic and clinical characteristics of pStage III patients stratified by ACT receipt
| Characteristic | ACT No (n = 428) |
ACT Yes (n = 1092) |
P value (Fisher or Wilcoxon) |
|
|---|---|---|---|---|
| Sex | Male | 233 (54.4%) | 572 (52.4%) | 0.4931 |
| Female | 195 (45.6%) | 520 (47.6%) | ||
| Age, years | Median (IQR) | 71 (62–77) | 66 (56–73) | < 0.0001 |
| Primary tumor location | Right (C-T) | 145 (33.9%) | 366 (33.5%) | 0.6755 |
| Left (D-RS) | 215 (50.2%) | 532 (48.7%) | ||
| Rectum (Ra, Rb) | 68 (15.9%) | 194 (17.8%) | ||
| Histological type | Differentiated | 410 (95.8%) | 1014 (92.9%) | 0.3993 |
|
Poorly differentiated/ Signet-ring cell |
15 (3.5%) | 50 (4.6%) | ||
| Other/Unknown | 3 (0.7%) | 28 (2.6%) | ||
| WES-defined ERBB2 amplification | Negative | 421 (98.4%) | 1070 (98.0%) | 0.8350 |
| Positive | 7 (1.6%) | 22 (2.0%) | ||
| RAS mutations | Wild type | 241 (56.3%) | 596 (54.6%) | 0.5201 |
| Mutant | 173 (40.4%) | 463 (42.4%) | ||
| Unknown | 14 (3.3%) | 33 (3.0%) | ||
| BRAF V600E mutation | No | 380 (88.8%) | 967 (88.6%) | 1.0000 |
| Yes | 34 (7.9%) | 89 (8.2%) | ||
| Unknown | 14 (3.3%) | 36 (3.3%) | ||
| MSI | MSI-high | 39 (9.1%) | 75 (6.9%) | 0.1608 |
| MSS | 373 (87.1%) | 960 (87.9%) | ||
| Unknown | 16 (3.7%) | 57 (5.2%) | ||
| Early post-operative MRD status (2–10 weeks post-surgery) | Negative | 337 (78.7%) | 696 (63.7%) | 0.0059 |
| Positive | 59 (13.8%) | 191 (17.5%) | ||
| Unknown | 32 (7.5%) | 205 (18.8%) |
IQR, interquartile range; Right (C-T), cecum to transverse colon (proximal colon); Left (D-RS), descending colon to Rectosigmoid (distal colon); Ra, upper rectum; Rb, lower rectum; MSI, microsatellite instability; MSS, microsatellite stable; MRD, molecular residual disease. P value was calculated using Fisher’s exact test among non-missing categories
Among patients without WES-defined ERBB2 amplification, ACT conferred a survival benefit (HR: 0.67; 95% CI: 0.54–0.83; P = 0.0002), with a 3 year DFS rate of 74.4% among those receiving ACT versus 67.3% in those without ACT (Fig. 4a). In contrast, no significant benefit from ACT was observed among patients with WES-defined ERBB2 amplification (HR: 1.13; 95% CI: 0.24–5.48; P = 0.8747), with 3 year DFS rates of 64.6% and 68.6%, respectively (Fig. 4b). To evaluate whether the association between ACT and DFS differed according to WES-defined ERBB2 amplification status after adjustment for other clinicopathological factors, we performed an interaction analysis including an ACT × WES-defined ERBB2 amplification interaction term (Fig. 4c). The interaction was not statistically significant (interaction HR: 1.94, 95% CI: 0.23–16.49, P = 0.5458).
Fig. 4.
Efficacy of ACT based on WES-defined ERBB2 amplification status.a, b Kaplan–Meier estimates of DFS stratified by ACT status among pStage III patients without WES-defined ERBB2 amplification (N = 1491) (a) and with WES-defined ERBB2 amplification (N = 29) (b). c Multivariable analysis of DFS including an ACT × WES-defined ERBB2 amplification interaction term among pStage III patients with complete clinicopathologic data (N = 1183). d, e Kaplan–Meier estimates of DFS stratified by ACT status in pStage III patients without WES-defined ERBB2 amplification and with early post-operative MRD-positivity (N = 223) (d), or early post-operative MRD-negativity (N = 1009) (e). Landmark analysis was performed for panels c-e. Analyses stratified by MRD status and ACT status among patients with WES-defined ERBB2 amplification are not included due to small sample size. DFS, disease-free survival; pStage, pathological stage; ACT, adjuvant chemotherapy
Given that early postoperative MRD status is previously shown to be strongly predictive of benefit of ACT, we further analyzed the efficacy of ACT among patients stratified by MRD status within 2–10 weeks post-surgery. Among patients without WES-defined ERBB2 amplification, those with MRD-positivity derived significant benefit from ACT (HR: 0.27, 95% CI: 0.18–0.39, P < 0.0001; 3 year DFS: 37.3% for ACT group vs. 0% for no ACT group; Fig. 4d). In contrast, no statistically significant benefit of ACT was observed among patients without WES-defined ERBB2 amplification who were MRD -negative (HR: 0.74, 95% CI: 0.53–1.02, P = 0.0674; 3 year DFS: 83.0% for ACT group vs. 78.1% for no ACT group; Fig. 4e). Analyses stratified by MRD and ACT status could not be performed among patients with WES-defined ERBB2 amplification due to low sample size.
In the WES-defined ERBB2 non-amplified subgroup, multivariable Cox analyses demonstrated a significant association between ACT and improved DFS, including in an additional model adjusted for age and early post-operative MRD status (Supplementary Fig. 3a, b). In contrast, in the WES-defined ERBB2 amplified subgroup, ACT was not significantly associated with improved DFS in a multivariable model adjusted for age and early post-operative MRD status (Supplementary Fig. 3c).
Prognostic value of ERBB2 copy number and MRD status in patients with WES-defined ERBB2 amplification
Next, we examined the prognostic significance of ERBB2 copy number and early post-operative MRD status among patients with WES-defined ERBB2 amplified resectable CRC. Patients who later developed a recurrence showed a trend towards higher ERBB2 copy numbers compared to those who remained recurrence-free, although the difference was not statistically significant (median copy number: 29.27 [IQR, 20.26–35.85] vs. 15.22 [IQR, 7.76–25.73], P = 0.0544) (Fig. 5a). In contrast, early post-operative MRD positivity was strongly associated with poor DFS (HR: 15.26, 95% CI: 3.94–59.20, P < 0.0001) (Fig. 5b). When analysis was limited to patients with pStage III disease, similar trends were observed with recurrent cases having higher ERBB2 copy numbers without statistical significance (Median Copy Number: 28.00 [IQR, 17.53–35.19] vs. 13.78 [IQR, 8.24–29.76], P = 0.2373), whereas MRD-positive patients exhibited significantly shorter DFS (HR: 9.15, 95% CI: 1.98–42.321, P = 0.0007) (Supplementary Fig. 4).
Fig. 5.
Prognostic significance and association betweenERBB2copy number and early post-operative MRD status (2–10 weeks post-surgery).a ERBB2 copy number stratified by recurrence status in patients with WES-defined ERBB2 amplification and pStages I-III CRC (N = 48). b Kaplan–Meier curve for DFS stratified by early post-operative MRD status in patients with WES-defined ERBB2 amplification and pStages I-III CRC (N = 40). cERBB2 copy number in patients with WES-defined ERBB2 amplification, stratified by early post-operative MRD status (N = 43). d Scatter plot of ERBB2 copy number and ctDNA levels (MTM/mL) at pre-operative (N = 41). Landmark analysis was performed for survival analyses in panel b. DFS, Disease-free survival
Finally, ERBB2 copy number did not significantly differ between patients who were MRD-positive or MRD-negative early post-operatively (median copy number : 27.07 [IQR, 21.32–35.19] vs. 16.86 [IQR, 8.36–31.27], P = 0.1781) (Fig. 5c) and showed no correlation with preoperative ctDNA levels (Pearson r = 0.1398, P = 0.3835; Spearman ρ = − 0.0176, P = 0.9131) (Fig. 5d).
Sensitivity analyses
Sensitivity analyses were conducted to assess whether the main findings were robust to changes in the definition of WES-defined ERBB2 amplification. The univariate association of WES-defined ERBB2 amplification with DFS showed an inconsistent trend below and above the p value 0.05 (Supplementary Fig. 5a). On the other hand, the multivariate analysis consistently showed WES-defined ERBB2 amplification to have no significant association with DFS at all cutoff values tested (Supplementary Fig. 5b).
In analyses evaluating efficacy of ACT among pStage III patients, all analyses remained consistent with those described above after changing the cutoff (Supplementary Fig. 5c–f).
Discussion
Given the low prevalence and limited existing literature, the clinical significance of HER2 amplification in resectable CRC has remained poorly defined. Leveraging WES data from the GALAXY study, this large-scale analysis characterized the clinicopathologic features of HER2-amplified resectable CRC, a rare but biologically distinct molecular subset. Our findings suggest that while WES-defined ERBB2 amplification is not significantly prognostic of DFS outcomes, a modest adverse prognostic effect cannot be excluded. Additionally, we found that early postoperative MRD positivity is a significant adverse prognostic factor in WES-defined ERBB2 amplified resectable CRC. Although the univariate analyses indicated a trend for association between WES-defined ERBB2 amplification and poorer DFS, its prognostic impact as a significant, independent adverse factor was not observed in the multivariate analysis. Recent studies have also reported that high HER2 expression is associated with poor prognosis in resectable CRC [13, 14]. However, a previous study reported no significant association between HER2 amplification and recurrences in resectable CRC [24], and the prognostic significance of HER2 amplification in resectable CRC therefore remains controversial and warrants further investigation.
We also examined whether WES-defined ERBB2 amplification influences responsiveness to ACT among patients with pStage III disease. Patients without WES-defined ERBB2 amplification derived a clear benefit from ACT. Interestingly, in this group without WES-defined ERBB2 amplification, early post-operative MRD -positivity (2–10 weeks post-surgery) further identified a subpopulation likely to benefit from ACT, whereas patients with MRD-negativity derived no significant benefit from ACT. In contrast, a clear benefit from ACT was not detected among patients with WES-defined ERBB2 amplification. However, because the ACT-by-ERBB2 amplification interaction was not statistically significant, the present data did not provide evidence that WES-defined ERBB2 amplification modifies the association between ACT and DFS. The limited number of amplified cases precluded a reliable assessment of whether WES-defined ERBB2 amplification modifies the association between ACT and DFS and validation in larger cohorts is required. Although HER2-mediated activation of the PI3K/AKT pathway is a plausible mechanism, as dysregulation of this pathway has been implicated in resistance to both 5-fluorouracil and oxaliplatin [25–30], our data did not allow conclusions regarding whether this mechanism influences ACT responsiveness in resectable CRC. The efficacy of HER2-targeted therapies is reported in mCRC [8–12] and ACT combined with HER2-targeted therapy is established as the standard of care in resectable HER2-positive breast cancer. Given the biological rationale for HER2 signaling and the uncertainty of our findings, larger prospective studies are warranted to evaluate whether HER2 amplification influences the association between ACT and DFS and whether HER2-targeted strategies have a role in the adjuvant setting for resectable CRC.
In this study, the absolute ERBB2 copy number was not significantly associated with recurrence or with early postoperative ctDNA-based MRD positivity and preoperative ctDNA levels. However, we found that early post-operative MRD positivity was a strong predictor of recurrence in WES-defined ERBB2 amplified cases, even upon limiting the analysis to pStage III disease. These results suggest that postoperative MRD status can aid recurrence risk stratification in WES-defined ERBB2 amplified CRC and may help identify candidates for adjuvant HER2-targeted therapy in future studies.
A strength of this study is the uniform WES-based assessment of ERBB2 copy number across all samples. HER2 amplification was defined with an ERBB2 copy number threshold of > 4.0 based on WES data. Nearly all (96%, 57/59) of WES-defined ERBB2 amplified tumors exceeded an ERBB2 copy number of > 5.0, a previously published threshold for HER2 positivity by IHC/FISH in mCRC [31]. Additionally, our observed frequency of WES-defined ERBB2 amplification (1.6%) and clinicopathological features of our cohort, such as higher HER2 amplification rates in pStage III, RAS/BRAF wild-type, and MSS tumors, are consistent with previous reports in resectable CRC [1, 5, 13, 14].
This study has several limitations. Its retrospective design introduces potential biases related to missing or incomplete data. Despite the large overall cohort, the small number of WES-defined ERBB2 amplified cases, limited number of events, and the relatively limited follow-up with evaluation restricted to 3 year DFS may have reduced statistical power. Therefore, the therapeutic implications of our findings should be interpreted cautiously and validated in larger datasets.
Conclusions
This study suggested that WES-defined ERBB2 amplification is not a significant adverse prognostic factor in resectable CRC after accounting for ctDNA-based MRD status and other clinicopathological variables and did not provide evidence that WES-defined ERBB2 amplification modifies the association between ACT and DFS. The limited number of amplified cases precluded a reliable assessment of whether WES-defined ERBB2 amplification modifies the association between ACT and DFS. To our knowledge, this is the largest study to date addressing WES-defined ERBB2 amplification as a prognostic and predictive biomarker in resected CRC. Further validation in larger-scale studies is warranted.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We thank all the patients and their families for their participation in GALAXY study as well as all study investigators and study teams for their contributions. CIRCULATE-Japan receives.
financial support from the Japan Agency for Medical Research and Development (grant 19ck0106447h0002-TY). The funders had no role in the study design; in data collection, analysis or interpretation; or in the writing of the report.
Author contributions
Conceptualization: H.B., T.H., A.H., K. Kishi, S.M., Y.N.; Data Curation: No contributions claimed.; Formal Analysis: A.F., A.H., K. Kishi, R.W.L.; Investigation: N.A., H.B., T.H., M.I., K.H., K. Kishi, K. Kato, D.K., M.K., S.M., Y.N., E.O., Y.S., H.T., I.T., J.W., M.Y., T.Y.; Methodology: A.F., A.H., K. Kishi, R.W.L.; Project Administration: A.H., K. Kishi, E.O., C.C.P., T.Y.; Software: A.F., A.H., K. Kishi, C.C.P.; Resources: N.A., H.B., T.H., M.I., K.H., K. Kishi, K. Kato, D.K., M.K., S.M., Y.N., E.O., Y.S., H.T., I.T., J.W., M.Y., T.Y.; Supervision: Y.D., H.E., E.O., M.U., T.Y.; Visualization: A.F., A.H., K. Kishi, C.C.P.; Writing – Original Draft: A.H., K. Kishi.; Writing – Review & Editing: N.A., H.B., Y.D., H.E., A.F., T.H., A.H., M.I., K.H., K. Kishi, K. Kato, T. Kato, D.K., M.K., R.W.L., S.M., Y.N., E.O., Y.S., H.T., I.T., M.U., M.Y., T.Y.
Funding
Open Access funding provided by The University of Osaka. K. Kishi, N.A., K.H., M.I., K. Kato, M.K., Y.N., T.H., H.T., M.U., and H.E. have nothing to disclose. A.F., C.C.P., and R.W.L. are employees of Natera, Inc., each receiving a salary and may own stock and/or stock options. A.H. reports payment or honoraria from Merck Biopharma Co., Ltd., Johnson & Johnson, Medtronic, and Fujifilm. Y.S. reports honoraria from Eli Lilly, Takeda Pharmaceutical Company Limited, Merck Biopharma, Taiho Pharmaceutical, Chugai Pharmaceutical, Sysmex, and Terumo. M.Y. reports payment or honoraria from Johnson & Johnson, Intuitive, Covidien Japan, Taiho Pharmaceutical, Takeda Pharmaceutical Company Limited, Eli Lilly, Kaken Pharmaceutical, and Merck Biopharma. H.B. reports honoraria from Ono Pharmaceutical, Taiho Pharmaceutical, and Guardant Health. Y.N. reports consulting fees from Guardant Health, Natera, Inc., Exact Sciences, Tempus AI, Inc., Pfizer Inc., and Shionogi & Co., Ltd., and honoraria from MSD K.K., Merck Biopharma, Daiichi Sankyo, Chugai Pharmaceutical, and Guardant Health. S.M. reports honoraria from Taiho Pharmaceutical, Chugai Pharmaceutical, Eli Lilly, and Takeda Pharmaceutical Company Limited, and advisory board participation for Exact Sciences. D.K. reports grants or contracts from Ono Pharmaceutical, MSD, Novartis, Servier, Janssen, IQVIA, Syneos Health, Cimic, and Cimicshiftzero, payment or honoraria from Takeda Pharmaceutical Company Limited, Chugai Pharmaceutical, Eli Lilly, MSD, Ono Pharmaceutical, Taiho Pharmaceutical, Bristol-Myers Squibb, Daiichi Sankyo, Pfizer Inc., Eisai, and Merck Biopharma, and advisory board participation for Takeda Pharmaceutical Company Limited and Seagen. J.W. reports institutional research support from Covidien Japan, Terumo, and Stryker Japan, and honoraria from Covidien Japan, Terumo, Eli Lilly, Takeda Pharmaceutical Company Limited, and Johnson & Johnson. I.T. reports grants or contracts from Medtronic, Johnson & Johnson, Intuitive, Medicaroid, Stryker, Anaut, Asensus, Olympus, TENMA Co, MIYARISAN, FUJIFILM, ASAHI, Applied Medical, Baxter, Kaken Pharmaceutical, Amco, and CONMED, consulting fees from Medtronic, Johnson & Johnson, Intuitive, Medicaroid, Eli Lilly, Stryker, Anaut, Asensus, and LivsMed, travel support from TENMA Co, and leadership roles with several professional societies. T. Kato reports honoraria from Chugai Pharmaceutical, Takeda Pharmaceutical Company Limited, Ono Pharmaceutical, Eli Lilly, and Asahi Kasei Pharma Corporation. Y.D. reports honoraria from Ono Pharmaceutical, Bristol-Myers Squibb, and MSD. E.O. reports honoraria from Eli Lilly, Takeda Pharmaceutical Company Limited, Ono Pharmaceutical, and Bristol-Myers Squibb. T.Y. reports institutional grants or contracts from Bristol-Myers Squibb, Caris MPI, Chugai Pharmaceutical, Daiichi Sankyo, Eisai, Exact Sciences, FALCO Biosystems, Medical & Biological Laboratories, Merus N.V., Miyarisan Pharmaceutical, Molecular Health GmbH, MSD, Natera, Inc., Nippon Boehringer Ingelheim, Ono Pharmaceutical, Pfizer Inc., Sysmex, Taiho Pharmaceutical, and Takeda Pharmaceutical Company Limited, consulting fees from Sumitomo Corporation and Indivumed, and honoraria from Chugai Pharmaceutical, Takeda Pharmaceutical Company Limited, Merck Biopharma, and Ono Pharmaceutical.
Data availability
Clinical and sequencing data from the GALAXY study are available under restricted access due to ethical and legal constraints. For detailed data sharing policies, please refer to the prior publications [17, 18]. Any requests will be reviewed within a time frame of 2 to 3 weeks by the CIRCULATE-Japan study steering committee to verify whether the request is subject to any intellectual property or confidentiality obligations. All data shared will be de-identified.
Code availability
The fully documented code to generate raw outputs from the R statistical computing environment for analyses related to this manuscript are deposited at the GitHub repository and can be accessed at https://github.com/Natera-TMED/Kishi-et-al_HER2_Galaxy.
Declarations
Ethical approval
This study was conducted in accordance with the Declaration of Helsinki. The Institutional Review Board of the National Cancer Center Japan approved of the overall analysis in this study.
Informed consent
All patients provided written informed consent before participating in the GALAXY study.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Singh H, Kang A, Bloudek L et al (2024) Systematic literature review and meta-analysis of HER2 amplification, overexpression, and positivity in colorectal cancer. JNCI Cancer Spectr. 10.1093/jncics/pkad082 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.The Cancer Genome Atlas Network (2012) Comprehensive molecular characterization of human colon and rectal cancer. Nature 487:330–337. 10.1038/nature11252 [DOI] [PMC free article] [PubMed]
- 3.Yoshikawa A, Nakamura Y (2022) Molecular basis of HER2-targeted therapy for HER2-positive colorectal cancer. Cancers (Basel). 10.3390/cancers15010183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Suwaidan AA, Lau DK, Chau I (2022) HER2 targeted therapy in colorectal cancer: New horizons. Cancer Treat Rev 105:102363. 10.1016/j.ctrv.2022.102363 [DOI] [PubMed] [Google Scholar]
- 5.Richman SD, Southward K, Chambers P et al (2016) HER2 overexpression and amplification as a potential therapeutic target in colorectal cancer: analysis of 3256 patients enrolled in the QUASAR, FOCUS and PICCOLO colorectal cancer trials. J Pathol 238:562–570. 10.1002/path.4679 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sawada K, Nakamura Y, Yamanaka T et al (2018) Prognostic and predictive value of HER2 amplification in patients with metastatic colorectal cancer. Clin Colorectal Cancer 17:198–205. 10.1016/j.clcc.2018.05.006 [DOI] [PubMed] [Google Scholar]
- 7.Bekaii-Saab TS, Lach K, Hsu LI et al (2023) Impact of anti-EGFR therapies on HER2-positive metastatic colorectal cancer: a systematic literature review and meta-analysis of clinical outcomes. Oncologist 28:885–893. 10.1093/oncolo/oyad200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tosi F, Sartore-Bianchi A, Lonardi S et al (2020) Long-term clinical outcome of trastuzumab and lapatinib for HER2-positive metastatic colorectal cancer. Clin Colorectal Cancer 19:256–262e252. 10.1016/j.clcc.2020.06.009 [DOI] [PubMed] [Google Scholar]
- 9.Nakamura Y, Okamoto W, Kato T et al (2021) Circulating tumor DNA-guided treatment with pertuzumab plus trastuzumab for HER2-amplified metastatic colorectal cancer: a phase 2 trial. Nat Med 27:1899–1903. 10.1038/s41591-021-01553-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Siena S, Di Bartolomeo M, Raghav K et al (2021) Trastuzumab deruxtecan (DS-8201) in patients with HER2-expressing metastatic colorectal cancer (DESTINY-CRC01): a multicentre, open-label, phase 2 trial. Lancet Oncol 22:779–789. 10.1016/s1470-2045(21)00086-3 [DOI] [PubMed] [Google Scholar]
- 11.Strickler JH, Cercek A, Siena S et al (2023) Tucatinib plus trastuzumab for chemotherapy-refractory, HER2-positive, RAS wild-type unresectable or metastatic colorectal cancer (MOUNTAINEER): a multicentre, open-label, phase 2 study. Lancet Oncol 24:496–508. 10.1016/s1470-2045(23)00150-x [DOI] [PubMed] [Google Scholar]
- 12.Sweeney CJ, Hainsworth JD, Bose R et al (2024) MyPathway human epidermal growth factor receptor 2 basket study: pertuzumab + trastuzumab treatment of a tissue-agnostic cohort of patients with human epidermal growth factor receptor 2-altered advanced solid tumors. J Clin Oncol 42:258–265. 10.1200/jco.22.02636 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pilati C, Soulabaille A, Gallois C et al (2025) ERBB2 comprehensive profiling and prognostication in stage iii colon cancer: findings from PETACC8 and IDEA-france cohorts. Gastroenterology 168:714–724e714. 10.1053/j.gastro.2024.10.046 [DOI] [PubMed] [Google Scholar]
- 14.Wu Z, Cheng Y, Wang H et al (2023) Distinct clinicopathological features and prognostic values of high-, low-, or non-expressing HER2 status in colorectal cancer. Cancers (Basel). 10.3390/cancers15020554 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Feng Y, Li Y, Huang D et al (2019) HER2 as a potential biomarker guiding adjuvant chemotherapy in stage II colorectal cancer. Eur J Surg Oncol 45:167–173. 10.1016/j.ejso.2018.10.059 [DOI] [PubMed] [Google Scholar]
- 16.Shimada Y, Yagi R, Kameyama H et al (2017) Utility of comprehensive genomic sequencing for detecting HER2-positive colorectal cancer. Hum Pathol 66:1–9. 10.1016/j.humpath.2017.02.004 [DOI] [PubMed] [Google Scholar]
- 17.Kotani D, Oki E, Nakamura Y et al (2023) Molecular residual disease and efficacy of adjuvant chemotherapy in patients with colorectal cancer. Nat Med 29:127–134. 10.1038/s41591-022-02115-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nakamura Y, Watanabe J, Akazawa N et al (2024) ctDNA-based molecular residual disease and survival in resectable colorectal cancer. Nat Med 30:3272–3283. 10.1038/s41591-024-03254-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Taniguchi H, Nakamura Y, Kotani D et al (2021) CIRCULATE-Japan: circulating tumor DNA-guided adaptive platform trials to refine adjuvant therapy for colorectal cancer. Cancer Sci 112:2915–2920. 10.1111/cas.14926 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Reinert T, Henriksen TV, Christensen E et al (2019) Analysis of plasma cell-free DNA by ultradeep sequencing in patients with stages I to III colorectal cancer. JAMA Oncol 5:1124–1131. 10.1001/jamaoncol.2019.0528 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Powles T, Assaf ZJ, Davarpanah N et al (2021) ctDNA guiding adjuvant immunotherapy in urothelial carcinoma. Nature 595:432–437. 10.1038/s41586-021-03642-9 [DOI] [PubMed] [Google Scholar]
- 22.Magbanua MJM, Swigart LB, Wu HT et al (2021) Circulating tumor DNA in neoadjuvant-treated breast cancer reflects response and survival. Ann Oncol 32:229–239. 10.1016/j.annonc.2020.11.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nakamura K, Aimono E, Oba J et al (2021) Estimating copy number using next-generation sequencing to determine ERBB2 amplification status. Med Oncol 38(4):36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Richman SD, Southward K, Chambers P et al (2016) HER2 overexpression and amplification as a potential therapeutic target in colorectal cancer: analysis of 3256 patients enrolled in the QUASAR, FOCUS and PICCOLO colorectal cancer trials. J Pathol 238:562–570. 10.1002/path.4679 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Venturini J, Massaro G, Lavacchi D et al (2024) The emerging HER2 landscape in colorectal cancer: The key to unveil the future treatment algorithm? Crit Rev Oncol Hematol 204:104515. 10.1016/j.critrevonc.2024.104515 [DOI] [PubMed] [Google Scholar]
- 26.Hsieh CC, Li TW, Li CC et al (2024) DKK1 as a chemoresistant protein modulates oxaliplatin responses in colorectal cancer. Oncogenesis 13:34. 10.1038/s41389-024-00537-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Huang JQ, Li HF, Zhu J et al (2021) SRPK1/AKT axis promotes oxaliplatin-induced anti-apoptosis via NF-κB activation in colon cancer. J Transl Med 19:280. 10.1186/s12967-021-02954-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yu T, An Q, Cao XL et al (2020) GOLPH3 inhibition reverses oxaliplatin resistance of colon cancer cells via suppression of PI3K/AKT/mTOR pathway. Life Sci 260:118294. 10.1016/j.lfs.2020.118294 [DOI] [PubMed] [Google Scholar]
- 29.Zhang Y, Liu X, Zhang J et al (2020) Inhibition of miR-19a partially reversed the resistance of colorectal cancer to oxaliplatin via PTEN/PI3K/AKT pathway. Aging 12:5640–5650. 10.18632/aging.102929 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Long S, Wang J, Weng F et al (2022) ECM1 regulates the resistance of colorectal cancer to 5-FU treatment by modulating apoptotic cell death and epithelial-mesenchymal transition induction. Front Pharmacol 13:1005915. 10.3389/fphar.2022.1005915 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fujii S, Magliocco AM, Kim J et al (2020) International harmonization of provisional diagnostic criteria for ERBB2-amplified metastatic colorectal cancer allowing for screening by next-generation sequencing panel. JCO Precis Oncol 4:6–19. 10.1200/po.19.00154 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Clinical and sequencing data from the GALAXY study are available under restricted access due to ethical and legal constraints. For detailed data sharing policies, please refer to the prior publications [17, 18]. Any requests will be reviewed within a time frame of 2 to 3 weeks by the CIRCULATE-Japan study steering committee to verify whether the request is subject to any intellectual property or confidentiality obligations. All data shared will be de-identified.
The fully documented code to generate raw outputs from the R statistical computing environment for analyses related to this manuscript are deposited at the GitHub repository and can be accessed at https://github.com/Natera-TMED/Kishi-et-al_HER2_Galaxy.