Factors associated with neoadjuvant therapy insensitivity and its prognostic impact in HER2-positive breast cancer

Abstract

Objective

A subset of patients with human epidermal growth factor receptor 2 positive (HER2+) breast cancer shows insensitivity to neoadjuvant therapy (NAT), often evidenced by imaging results indicating stable disease (SD) or progressive disease (PD), which may reflect intrinsic resistance to treatment. We aimed to investigate the factors associated with NAT insensitivity and its prognostic value in HER2+ breast cancer.

Methods

This study included consecutive patients with HER2+ breast cancer who received NAT consisting of chemotherapy combined with anti-HER2 monoclonal antibodies. NAT insensitivity was defined as SD or PD on the basis of treatment response evaluations. Statistical analyses were conducted on the collected clinical data, and HER2 heterogeneity was subsequently assessed.

Results

A total of 541 patients were included in the study, among whom 63 (11.6%) were categorized as NAT-insensitive group and 478 (88.4%) as NAT-sensitive group. Hormone receptor (HR) status (P=0.033), HER2 status (P=0.036) and anti-HER2 therapy (P=0.007) were associated with NAT sensitivity. NAT-insensitive group had a significantly shorter event-free survival (EFS) (3-year: 69.4% vs. 94.3%; P<0.001) and remained an independent prognostic factor according to Cox models [hazard ratio (HR)=8.637; 95% confidence interval (95% CI), 3.091−24.136; P<0.001]. Exploratory analysis revealed a greater proportion of HER2 heterogeneity in the NAT-insensitive group (19.4% vs. 4.3%; P=0.035).

Conclusions

HR positivity, HER2 2+/fluorescence in situ hybridization (FISH)+ status, and trastuzumab monotherapy are associated with NAT insensitivity, and NAT insensitivity independently indicates poor EFS. This study also highlights the need for prospective studies to clarify the role of HER2 heterogeneity and other mechanisms involved in predicting the response to NAT.

Introduction

Human epidermal growth factor receptor 2 positive (HER2+) breast cancer is a distinct molecular subtype that has a poor prognosis but is sensitive to anti-HER2 targeted therapies (1). Neoadjuvant therapy (NAT) has become an important part of the comprehensive management of HER2+ breast cancer. The primary objective of NAT is to downsize tumors preoperatively through chemotherapy and targeted therapies, thereby increasing the success rate of surgery and serving as an “in vivo experiment” to evaluate the efficacy of drugs (2). For patients with HER2+ breast cancer, NAT is particularly important because it can significantly increase the pathological complete response (pCR) rate and improve long-term prognosis. In addition, the non-pCR population can be treated with intensified therapy in the adjuvant setting. Although pCR is currently the gold standard, it is prognostic but cannot be predicted. Image-based evaluation remains routine in the clinic to guide the NAT treatment process (3,4).

The selection of NAT regimens for HER2+ breast cancer has been guided by extensive evaluation in several pivotal studies. The NOAH trial was a landmark study demonstrating that, compared with chemotherapy alone, chemotherapy combined with trastuzumab doubled the tpCR rate (38% vs. 19%, P=0.001). Additionally, this combination increased 3-year event-free survival (EFS) (71% vs. 56%), marking the beginning of targeted NAT for HER2+ breast cancer (5). In addition to NOAH, studies such as NeoSphere and PEONY further demonstrated that compared with trastuzumab monotherapy, dual blockade with pertuzumab and trastuzumab significantly increased the pCR rate (6,7). These trials solidified dual blockade as a key approach in the neoadjuvant treatment of early-stage HER2+ breast cancer. In the KRISTINE trial, the combination of chemotherapy, trastuzumab, and pertuzumab (TCbH+P) resulted in a superior pCR rate compared with trastuzumab emtansine (T-DM1) with pertuzumab (56% vs. 44%, P=0.016). Long-term follow-up revealed that patients who achieved pCR with TCbH+P had a 3-year invasive disease-free survival (IDFS) rate of 97.5%, whereas those without pCR had a 3-year IDFS of 84.2% (8,9). Together, these studies have established the combination of chemotherapy, trastuzumab, and pertuzumab as the standard neoadjuvant regimen for patients with stage II−III HER2+ breast cancer (10).

Although major progress has been made in increasing the survival of patients with HER2+ breast cancer, the addition of anti-HER2 targeted therapies has not resulted in optimal outcomes for all patients. Some patients exhibit NAT insensitivity, often evidenced by imaging results indicating stable disease (SD) or progressive disease (PD), suggesting the potential for primary resistance (11).

In this study, we defined SD or PD as a lack of sensitivity to NAT (NAT insensitive), whereas partial response (PR) or complete response (CR) was considered NAT sensitive. We aimed to investigate the factors associated with NAT insensitivity in HER2+ breast cancer and its prognostic impact.

Materials and methods

Study design

This was a multicenter retrospective cohort study conducted at the Cancer Hospital Chinese Academy of Medical Sciences and Sanhuan Cancer Hospital between January 1, 2017, and December 31, 2023. All eligible patients were consecutively included during the study period. The study was approved by the Institutional Review Board of the National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital (Approval No. 12-123/657).

Patients and eligibility

This study included 639 consecutive patients with HER2+ breast cancer who received NAT. Inclusion criteria: 1) Histologically confirmed HER2+ breast cancer; 2) initial NAT consisting of chemotherapy combined with monoclonal antibodies, such as trastuzumab, trastuzumab + pertuzumab, or inetetamab + pertuzumab; and 3) availability of important NAT data. Exclusion criteria: 1) Evidence or suspicion of distant metastasis on imaging prior to treatment; 2) treatment regimens without anti-HER2 therapy; 3) use of other anti-HER2 therapies, such as antibody-drug conjugates (ADCs) or tyrosine kinase inhibitors (TKIs), in the initial regimen; 4) inability to assess treatment efficacy during the NAT period; or 5) incomplete medical records, preventing the retrieval of key information.

Definitions

In this study, insensitivity to NAT was defined as the absence of a PR or CR to anti-HER2 monoclonal antibody therapy or the presence of tumor progression or metastasis during treatment (12). EFS is defined as the time from the initiation of NAT to the time of disease progression that prevents surgery or the first documented recurrence of disease, whether local, regional, or distant, as well as death. pCR was defined as the absence of invasive disease in both the breast and axillary lymph nodes, as determined by the pathological report following surgery.

Data collection

Baseline clinicopathologic variables and treatment details were extracted at baseline and at each treatment cycle from electronic medical records and the imaging system. Treatment responses were initially evaluated and documented in medical records by the attending physicians during routine clinical practice. During data collection, the first author rechecked and verified the treatment responses against the original records, and discrepancies were adjudicated by a senior oncologist. Double data entry was performed by two researchers to ensure accuracy. Conflicting entries were resolved via source-document review. Records with unretrievable essential data were excluded. Follow-up information was obtained from outpatient visits, inpatient records, and telephone contact when necessary. Survival data were censored at the last follow-up (December 31, 2024). The median follow-up duration was 36 (range: 28−44) months.

Assessment of HER2 heterogeneity

In this study, HER2 heterogeneity was assessed using pretreatment paraffin-embedded core biopsy samples. On the basis of the intensity of HER2 protein staining by immunohistochemistry (IHC), tumor cells were classified as 1+, 2+, or 3+. Cells with HER2 2+ or 3+ staining were considered positive cells and were collectively used for categorization. HER2 staining patterns are divided into three categories: focal (10%−29% positive cells), variable (30%−79%), and homogeneous (≥80%) (13). Tumors exhibiting focal or variable staining patterns were defined as having HER2 heterogeneity.

For clinical HER2 determination and study eligibility, the American Society of Clinical Oncology/College of American Pathologists guidelines were strictly followed [IHC 3+ or IHC 2+/fluorescence in situ hybridization (FISH)+]. The IHC 2+/3+ cell-based categorization described here was used exclusively to describe intra-tumoral expression patterns for heterogeneity research purposes, not to assign clinical HER2 status.

Statistical analysis

Comparisons between the NAT-sensitive and NAT-insensitive groups were made using appropriate statistical tests: independent-sample t tests or the Wilcoxon rank-sum test for continuous variables and Chi-square tests for categorical data. To explore factors associated with NAT insensitivity, univariate and multivariate logistic regression analyses were employed. Analyses were performed based on available-case data, and no imputation was applied.

For survival analysis, EFS was calculated from the initiation of NAT. Propensity score matching (PSM) was applied to balance potential confounders between groups. Propensity scores were estimated using a logistic regression model, with clinical stage, T stage, hormone receptor (HR) status, HER2 status, and anti-HER2 therapy included as covariates because of their potential influence on outcomes. A 1:2 nearest-neighbour matching without replacement was applied. After matching, the balance of baseline characteristics between groups was assessed using standardized mean difference (SMD), with an SMD<0.10 considered adequate balance. Survival probabilities were assessed using the Kaplan-Meier method. Cox regression analysis was performed to identify factors for prognosis. All the statistical analyses were performed using SPSS software (Version 27.0; IBM Corp., New York, USA) and R software (Version 4.4.1; R Foundation for Statistical Computing, Vienna, Austria), with statistical significance defined as P<0.05.

Results

Patient inclusion

A total of 639 patients who received NAT between 2017 and 2023 were identified using medical records and imaging systems. After the inclusion and exclusion criteria were applied, 98 patients were excluded. Ultimately, 541 patients were included in the study, as shown in the flowchart (Figure 1). Among these patients, 63 (11.6%) patients were classified as NAT-insensitive group, whereas 478 (88.4%) patients were classified as NAT-sensitive group.

Figure 1.

Flowchart of patient inclusion and grouping. HER2, human epidermal growth factor receptor 2; TKI, tyrosine kinase inhibitor; ADC, antibody-drug conjugate.

Clinical characteristics and initial treatment regimens

The clinical characteristics and treatment details of the patients are summarized in Table 1. Statistical analysis using independent sample tests revealed that the NAT-insensitive group had greater proportions of patients with T4 stage disease (17.5% vs. 6.3%, P=0.022), clinical stage III disease (79.4% vs. 71.3%, P=0.018), HR+ (71.4% vs. 57.2%, P=0.025), HER2 2+/FISH+ status (23.8% vs. 13.6%, P=0.039) and trastuzumab monotherapy (41.3% vs. 24.9%, P=0.008) (Table 1).

Table 1. Clinical characteristics and treatment details of two groups.

Characteristic	n (%)		P
	NAT-insensitive group (N=63)	NAT-sensitive group (N=478)
HR, hormone receptor; HER2, human epidermal growth factor receptor 2; FISH, fluorescence in situ hybridization; NAT, neoadjuvant therapy.
Age at diagnosis (year)			0.373
<35	12 (19.0)	59 (12.3)
35−50	28 (44.4)	206 (43.1)
51−60	18 (28.6)	153 (32.0)
>60	5 (7.9)	60 (12.6)
cT stage			0.022
0/1	5 (7.9)	44 (9.3)
2	38 (60.3)	320 (67.9)
3	9 (14.3)	77 (16.3)
4	11 (17.5)	30 (6.3)
cN stage			0.203
0	7 (11.1)	78 (16.4)
1	7 (11.1)	74 (15.5)
2	34 (54.0)	191 (40.1)
3	15 (23.8)	133 (27.9)
Clinical stage			0.018
II	13 (20.6)	135 (28.7)
III	50 (79.4)	335 (71.3)
Histological grade			0.836
2	28 (68.3)	200 (66.7)
3	13 (31.7)	100 (33.3)
HR			0.025
Negative	18 (28.6)	204 (42.8)
Positive	45 (71.4)	273 (57.2)
HER2			0.039
2+/FISH+	15 (23.8)	65 (13.6)
3+	48 (76.2)	413 (86.4)
Chemotherapy			0.135
Taxane	56 (88.9)	446 (93.9)
Anthracycline-taxanes	7 (11.1)	29 (6.1)
Anti-HER2 therapy			0.008
Monotherapy	26 (41.3)	119 (24.9)
Dual therapy	37 (58.7)	359 (75.1)

Treatment patterns in NAT-insensitive group

The neoadjuvant treatment process for the 63 patients in NAT-insensitive group is further detailed in Supplementary Table S1. In NAT-insensitive group, treatment modification was common because of poor response, which occurred between cycles 2 and 6, and involved regimen switching or proceeding directly to surgery. Among patients who switched to alternative regimens, the median number of additional cycles was 3 (range: 2−6). Some patients extended their NAT up to 8−10 cycles after switching regimens, whereas others did not change the regimen but did prolong the treatment duration.

With respect to treatment response, 63.5% (40/63) of patients experienced SD or SD with slight improvement (SDa) after the initial treatment, whereas 30.2% (19/63) experienced SD with tumor enlargement (SDb). Four (6.3%) patients demonstrated PD during NAT (Supplementary Table S1). Postoperative pathology revealed that only 3 (4.8%) patients achieved pCR, whereas the majority had residual disease. Approximately half of the patients received adjuvant chemotherapy, and nearly all received postoperative anti-HER2 targeted therapy. The details of the postoperative treatments are summarized in Supplementary Table S2.

Factors associated with NAT insensitivity

To identify factors associated with insensitivity to NAT, variables with significant differences in independent-sample tests were included in logistic regression analyses, as detailed in Table 2. Univariate logistic regression analysis revealed the following factors associated with NAT insensitivity: HR+ status [odds ratio (OR)=1.868; P=0.033], HER2 2+/FISH+ status (OR=1.976; P=0.036), and trastuzumab monotherapy (OR=2.114; P=0.007). Variables with P<0.2 in the univariate analysis were further analyzed using multivariate logistic regression. The results revealed that only trastuzumab monotherapy was significantly associated with NAT insensitivity (OR=2.006, P=0.014).

Table 2. Univariate and multivariate logistic analyses to identify factors associated with NAT insensitivity.

Factor	Univariate analysis			Multivariate analysis
	OR (95% CI)	P		OR (95% CI)	P
NAT, neoadjuvant therapy; HR, hormone receptor; HER2, human epidermal growth factor receptor 2; OR, odds ratio; 95% CI, 95% confidence interval.
T3/T4 vs. T0/T1/T2	1.706 (0.968−3.005)	0.065		1.659 (0.906−3.038)	0.101
Stage: III vs. II	1.708 (0.883−3.303)	0.112		1.420 (0.706−2.855)	0.325
HR+ vs. HR−	1.868 (1.050−3.323)	0.033		1.677 (0.924−3.403)	0.089
HER2 2+/FISH+ vs. HER2 3+	1.976 (1.046−3.733)	0.036		1.797 (0.915−3.530)	0.089
Monotherapy vs. dual-therapy	2.114 (1.229−3.638)	0.007		2.006 (1.153−3.489)	0.014

Impact of NAT insensitivity on survival

To analyze the prognostic value of NAT insensitivity, PSM was performed to balance potential confounding factors between the NAT-insensitive group and the NAT-sensitive group. PSM (1:2) was performed, and the clinical stage, T stage, HR status, HER2 status, and anti-HER2 therapy status were balanced, with SMD<0.10, as shown in Supplementary Table S3. Survival analysis was then conducted with a median follow-up period of 36 (range: 28−44) months. The EFS Kaplan-Meier curve categorized by NAT insensitivity is shown in Figure 2. Kaplan-Meier analysis revealed that compared with the NAT-sensitive group, the NAT-insensitive group had a significantly shorter EFS (P<0.001), with 3-year EFS rates of 69.4% and 94.3%, respectively. We subsequently included the potential prognostic factors and conducted both univariate and multivariate Cox regression analyses. As shown in Table 3, among the various clinical factors, only sensitivity to NAT was identified as an independent prognostic factor (HR=8.637; 95% CI, 3.091−24.136; P<0.001).

Figure 2.

EFS of HER2-positive patients based on their response to NAT. EFS, event-free survival; HER2, human epidermal growth factor receptor 2; NAT, neoadjuvant therapy.

Table 3. Univariate and multivariate Cox regression analyses to identify prognostic factors for EFS.

Factor	Univariate analysis			Multivariate analysis
	HR (95% CI)	P		HR (95% CI)	P
EFS, event-free survival; HR, hormone receptor; HER2, human epidermal growth factor receptor 2; HR, hazard ratio; 95% CI, 95% confidence interval.
T3/T4 vs. T0/T1/T2	1.060 (0.458−2.453)	0.892		1.116 (0.465−2.681)	0.805
Stage: III vs. II	4.349 (0.585−31.312)	0.151		5.167 (0.673−39.653)	0.114
HR+ vs. HR−	0.619 (0.268−1.432)	0.263		0.593 (0.246−1.431)	0.245
HER2 2+/FISH+ vs. HER2 3+	1.342 (0.457−3.947)	0.593		0.730 (0.236−2.260)	0.585
Monotherapy vs. dual-therapy	1.062 (0.463−2.437)	0.886		1.533 (0.648−3.630)	0.331
Insensitivity vs. sensitivity	6.957 (2.580−18.756)	<0.001		8.637 (3.091−24.136)	<0.001

HER2 heterogeneity is associated with NAT insensitivity

To investigate potential contributors to NAT insensitivity, we evaluated intra-tumoral HER2 heterogeneity by analyzing HER2 staining patterns in pretreatment core biopsy samples. All the evaluable samples from the NAT-sensitive group and the matched NAT-insensitive group were included. After validation, no significant differences were observed in clinical characteristics between the two subsets, as shown in Supplementary Table S4. The staining patterns of HER2 protein, which were divided into three categories, namely, focal (10%−29% positive cells), variable (30%−79%), and homogeneous (≥80%), are shown in Figure 3. Samples with focal or variable HER2 staining patterns were classified as having intra-tumoral HER2 heterogeneity. In the NAT-insensitive group, 36 pre-treatment biopsy samples were available for HER2 heterogeneity assessment. Among them, 3 showed focal expression, 4 exhibited variable expression, and 29 demonstrated homogeneous HER2 staining. In the matched NAT-sensitive group, 46 evaluable samples were analyzed, with 1 case showing focal expression, 1 showing variable expression, and 44 demonstrating homogeneous staining. As shown in Figure 4, the proportion of HER2 heterogeneity (focal or variable staining) was significantly higher in the NAT-insensitive group than in the NAT-sensitive group (19.4% vs. 4.3%, P=0.035).

Figure 3.

HER2 protein staining intensity and three staining patterns. (A) Focal; (B) Variable; (C) Homogeneous. (×10). HER2, human epidermal growth factor receptor 2.

Figure 4.

Proportion of HER2 heterogeneity in NAT-insensitive and NAT-sensitive groups. HER2, human epidermal growth factor receptor 2; NAT, neoadjuvant therapy.

Discussion

Although many patients with HER2+ breast cancer exhibit excellent responses to NAT, approximately 10%−20% of patients are insensitive, suggesting a distinctive biological entity (6,14). Factors associated with NAT insensitivity in HER2+ breast cancer and its prognostic impact have rarely been reported. In this real-world study, we investigated the clinical and pathological characteristics and prognostic implications of NAT insensitivity in patients with HER2+ breast cancer.

In this study, the incidence of insensitivity to NAT with anti-HER2 monoclonal antibody therapy was 11.6%. Significant differences in T stage, clinical stage, HR status, and HER2 status were detected between the NAT-sensitive and NAT-insensitive groups. In the NAT-insensitive group, a greater proportion of patients presented with T4 tumors involving skin invasion, suggesting that tumor size is associated with treatment response. These findings align with those of previous research that reported that the probability of achieving pCR decreases significantly in tumors larger than 5 cm (15). As the tumor stage increases, tumor burden increases, potentially altering tumor microenvironment. Changes such as increased angiogenesis and larger hypoxic regions may affect drug delivery and distribution, indirectly contributing to reduced drug efficacy and resistance (16-18). One of the goals of NAT is to downstage locally advanced tumors, enabling breast-conserving surgery or lumpectomy with radiation instead of mastectomy. However, our results revealed a greater proportion of patients with stage III disease (locally advanced) in the NAT-insensitive group, suggesting that patients with locally advanced disease may be more prone to treatment resistance, complicating efforts to achieve down-staging.

In this study, the proportion of HER2+/HR+ patients in the NAT-insensitive group was 71.4%, which was significantly greater than that in the NAT-sensitive group (57.2%). HER2+/HR+ breast cancer represents a distinct subtype, with bidirectional crosstalk between the estrogen receptor and HER2 pathways, influencing the response to both HER2-targeted therapies and endocrine therapy (ET). A retrospective analysis indicated that HR positivity is associated with a greater risk of poor response to NAT, as reflected in the postoperative residual cancer burden scores (19). In clinical trials involving trastuzumab or trastuzumab plus pertuzumab, HR+ patients had lower pCR rates than HR− patients did (20). Combined targeting strategies that include ET alongside anti-HER2 treatment have been tested in several trials, with the addition of ET leading to an increase in pCR rates; however, these differences did not reach statistical significance (21-23). Further studies are needed to identify the optimal treatment strategies for HER2+/HR+ breast cancer, particularly in patients who develop resistance during treatment. In this study, three patients in the NAT-insensitive group received a combination of ET and anti-HER2 treatment after an inadequate initial response. However, owing to the small sample size, further analysis was not possible.

Multiple studies have demonstrated that higher levels of HER2 protein expression are positively linked to tumor reliance on HER2 signalling. Among early-stage breast cancer cases, the strongest therapeutic response is typically observed in tumors with an IHC HER2 score of 3+, followed by those with a score of 2+ accompanied by HER2 gene amplification (24). Notably, HER2 3+ breast cancer has a higher pCR rate than HER2 2+/FISH+ breast cancer does (P≤0.05) (25). Consistent with these findings, our study suggested that the proportion of HER2 2+ patients was greater in the NAT-insensitive group. However, a previous study revealed that the prognosis for HER2 2+/FISH+ tumors is similar to that for HER2 3+ tumors (26). Additionally, different patterns of HER2 expression may be present within the tumor, suggesting that HER2 heterogeneity could impact treatment response (27).

Notably, both univariate and multivariate logistic regression analyses indicated that the likelihood of treatment insensitivity was greater with trastuzumab monotherapy than with dual-HER2 blockade therapy. Mechanisms contributing to trastuzumab resistance include reduced binding affinity to HER2, activation of alternative or downstream signalling pathways and evasion of antibody-dependent cell-mediated cytotoxicity, which have been well documented in numerous preclinical studies and clinical trials (28,29). To overcome these challenges, various complementary strategies have been developed, with the combination of trastuzumab and pertuzumab showing significant efficacy. Consequently, the combination of taxanes, carboplatin, and dual-HER2 blockade has become the preferred regimen for neoadjuvant treatment of HER2+ breast cancer. Moreover, recent clinical trials have investigated the effectiveness of adding small-molecule TKIs to trastuzumab, as well as regimens such as T-DM1 combined with pertuzumab and monotherapy with trastuzumab deruxtecan (T-Dxd), in the neoadjuvant setting. Studies have demonstrated that such combinations significantly increase the pCR rate (30-32). A dedicated study, NeoPaTHer, designed a prospective clinical trial to investigate the efficacy and safety of adding pyrotinib to chemotherapy combined with trastuzumab in patients who showed no response after two cycles of treatment. The results demonstrated that the addition of pyrotinib increased the pCR rate by 1.9 times (33). These findings align with the results of our study, in which 6 out of 10 patients who responded to treatment regimen 2 received a regimen that included pyrotinib. Notably, in the NeoPaTHer study, treatment regimens were altered after two cycles for patients who were unresponsive, which may not be the optimal time to assess drug efficacy. Furthermore, because the study was conducted earlier, the initial treatment used was a combination of trastuzumab and chemotherapy. Large-scale clinical trials are needed to explore the optimal treatment strategies for patients who develop resistance to chemo + HP regimens, including the use of pyrotinib, ADCs, and other emerging therapies.

After analyzing potential factors associated with NAT insensitivity, we investigated the impact of insensitivity itself on patient survival. To eliminate confounding factors, PSM was performed. Survival analysis revealed that when clinical and pathological factors and anti-HER2 therapy were controlled, the 3-year EFS rate for the NAT-insensitive group was significantly lower than that for the sensitive group (69.4% vs. 94.3%). Cox regression analysis revealed that NAT insensitivity is an independent prognostic factor for poor outcomes. These findings highlight that treatment sensitivity is a crucial factor influencing patient prognosis. Moreover, these observations raise the possibility that intrinsic biological factors may drive the development of resistance.

HER2 intra-tumoral heterogeneity refers to variations in HER2 expression or gene amplification within different regions of the same tumor. In breast cancer, HER2 heterogeneity may influence the response to HER2-targeted therapies. A recent clinical trial involving T-DM1 plus pertuzumab suggested that HER2 heterogeneity may negatively impact the pCR rate in patients with breast cancer; however, that study assessed heterogeneity using FISH, a method that is costly and less feasible in routine clinical practice (27). Another study in advanced HER2+ breast cancer also demonstrated that focal or variable HER2 staining was associated with shorter median progress-free survival (PFS) (34). Despite its potential significance, intra-tumoral HER2 heterogeneity is not routinely evaluated in clinical settings. Nonetheless, both previous research and the findings from our study indicate that HER2 heterogeneity may be associated with treatment insensitivity and poor prognosis. These results underscore the potential clinical value of assessing HER2 heterogeneity prior to therapy. Such evaluation could help identify tumors less likely to respond to standard regimens and guide the development of personalized treatment strategies—such as incorporating agents with bystander effects to eliminate HER2-negative tumor subclones. However, the clinical utility of this approach requires validation in prospective clinical trials.

This study has several limitations. The relatively small sample size in the NAT-insensitive group restricted our ability to thoroughly analyze factors associated with treatment sensitivity. Additionally, this study did not conduct an in-depth molecular analysis of the mechanisms underlying tumor resistance, which should be addressed in future research.

Conclusions

HR positivity, HER2 2+/FISH+ status, and trastuzumab monotherapy are associated with NAT insensitivity. NAT-insensitive patients demonstrated poorer EFS. Intra-tumoral HER2 heterogeneity is associated with patients’ sensitivity to neoadjuvant anti-HER2 therapy. Further investigations are warranted to identify the underlying mechanisms involved. More large-scale clinical trials are needed to explore the optimal treatment strategies for these patients.

SUPPLEMENTARY DATA

Supplementary data to this article can be found online.

Funding Statement

This work was supported by the CAMS Innovation Fund for Medical Sciences (CIFMS, No. 2021-I2M-1-014, 2023-I2M-C&T-B-077 and 2023-I2M-C&T-B-081), the Beijing Natural Science Foundation (No. L258033) and the National High Level Hospital Clinical Research Funding (No. 2025-LYZX-D-A02).

Author contributions

Data curation, formal analysis and writing – original draft: YH Han and B Lan; Data curation: ZX Peng, YR Jia, DY Ji, XZ Tian and YT Yang; Data curation and investigation: D Sang, JY Wang, F Ma, Y Luo, and P Zhang; Methodology: XM Xue; Conceptualization, writing – review & editing: BH Xu and Y Fan. All the authors contributed to the manuscript and approved the final submitted version.