Comparison of the Response to Neoadjuvant Therapy Between Immunohistochemistry HER2 (3+) and HER2 (2+)/ISH+ Early-Stage Breast Cancer: A Retrospective Multicenter Cohort Study

Abstract

Background

According to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) criteria, both immunohistochemical HER2 (3+) and HER2 (2+)/in situ hybridization (ISH) amplified [HER2 (2+)/ISH+] breast cancers (BCs) fall under the HER2-positive BC category. However, there is a lack of studies exploring the difference of neoadjuvant therapeutic response between patients with HER2 (3+) and HER2 (2+)/ISH+ early BC. We aimed to evaluate the neoadjuvant therapeutic response, long-term outcome, and intrinsic subtype heterogeneity between HER2 (3+) and HER2 (2+)/ISH+ BC.

Methods

We examined 2 distinct cohorts. Cohort 1 (C1) encompassed 2648 patients with HER2-positive early BC diagnoses, and they received neoadjuvant therapy (NT) and surgery between January 1, 2009 and December 31, 2022, from the Shanghai Jiao Tong University Breast Cancer Data Base. Cohort 2 (C2) comprised 135 patients with early-stage HER2-positive BC who underwent NT and surgery at Henan Cancer Hospital from January 1, 2021, to December 31, 2022. These patients had available genomic and transcriptomic data at their disposal. C1 and C2 were further categorized into 2 patient cohorts as follows: (1) patients with IHC HER2 (3+) early BC [HER2 (3+) group], (2) patients with HER2 (2+)/ISH+ early BC [HER2 (2+)/ISH+ group]. Among those excluded from the analysis were patients < 18 years or >80 years of age. Clinicopathological parameters, long-term outcomes, and intrinsic subtypes were analyzed.

Results

In the C1 population, 83.7% had HER2 (3+) BC, while 16.3% had HER2 (2+)/ISH+ BC. Patients with HER2 (3+) had a significantly higher pathological complete response (PCR) rate (38.9%) than patients with HER2 (2+)/ISH+ (18.1%; P < .001), but the disease-free survival (DFS) was comparable after a median follow-up of 29 months (P = .556). The addition of trastuzumab or trastuzumab plus pertuzumab to neoadjuvant chemotherapy (NAC) improved PCR rates and DFS in HER2 (3+) BC but not in HER2 (2+)/ISH+ BC. In the C2 population, 97.75% HER2 (3+) and 52.17% HER2 (2+)/ISH+ were HER2 enriched (HER2E) subtype (P < .001). HER2E showed increased PCR rates compared to non-HER2E (P = .004).

Conclusions

Compared to HER2 (3+) BC, the limited effectiveness of neoadjuvant trastuzumab and pertuzumab therapy for HER2 (2+)/ISH+ BC is due to subtype heterogeneity. Reassessment of targeted therapy efficacy in patients with HER2 (2+)/ISH+ BC is essential.

This cohort investigation provides compelling evidence of a significant difference in the response to neoadjuvant therapy between patients with HER2 (3+) and HER2 (2+)/ISH+.

Implications for Practice.

This cohort investigation provides compelling evidence of a significant difference in the response to neoadjuvant therapy between patients with HER2 (3+) and HER2 (2+)/ISH+. Patients with HER2 (2+)/ISH+ encounter challenges in deriving benefit from neoadjuvant targeted therapy, potentially due to intrinsic subtype differences. Based on our findings, we recommend considering HER2 protein expression when making neoadjuvant targeted treatment recommendations for patients with early-stage HER2-positive breast cancer. Reevaluation of the value of targeted therapy in patients with HER2 (2+)/ISH+ is necessary.

Introduction

HER2-positive breast cancer (BC) typically features an overexpression or amplification of the human epidermal growth factor receptor 2 (HER2) gene,^1-4 and makes up approximately 15%-20% of all BC incidences. It is also strongly correlated with aggressive tumor development and extremely worse patient outcome, when left untreated.^5-7 According to the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) criteria,^2,4,8 both immunohistochemistry-detected HER2 (3+) and HER2 (2+)/in situ hybridization (ISH) amplified [HER2 (2+)/ISH+] BCs fall under the HER2-positive BC category.

At present, anti-HER2 therapy plus chemotherapy has emerged as a highly efficacious and critical intervention for patients with early invasive stage^9-12 or advanced BC^13-16. HER2 (3+) and HER2 (2+)/ISH+ BCs are equally considered candidates for anti-HER2 therapy.

Neoadjuvant therapy (NT) is a recommended intervention for patients with early-stage HER2-positive BC, and it facilitates tumor downstaging, enhances breast-conserving surgery rates, and improves treatment response. Moreover, trastuzumab and pertuzumab introduction to neoadjuvant chemotherapy (NAC) further augments pathological complete response (PCR)^17-19, which is closely related to increased disease-free survival (DFS) and overall survival (OS) among patients with HER2-positive BC.²⁰ Despite the well-established NT efficiency among patients with HER2-positive BC, there is still 20%-60% patients who fail to achieve PCR after HER2-targeted therapy, with or without NAC. Hence, there is a substantial need for elucidation of factors that influence NT response in this particular subtype.

Prediction analysis microarray 50 (PAM50), which was developed using microarray and quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) data, has been successfully introduced reliable prognostic and indicator information to the standard parameters for patients with BC.²¹ Moreover, in NT, intrinsic HER2-enriched (HER2E) patients defined with PAM50 exhibit a higher PCR rate, along with a considerably worse relapse-free survival outcome, relative to non-HER2E patients.^22,23

A small sample retrospective study has demonstrated that among patients with early HER2-positive BC receiving NAC combined with trastuzumab and pertuzumab, the PCR rate was 55.1% for patients with HER2 (3+), significantly higher than the 17.6% rate observed in patients with HER2 (2+)/ISH+ (P < .001).²⁴ However, the value of introducing targeted therapies, namely, trastuzumab and pertuzumab, to NAC in the aforementioned 2 cohorts is not well-defined.

Given the limitations of our current understanding of this particular field, herein, we compared the NT response, long-term patient outcome, and intrinsic subtype heterogeneity between HER2 (3+) and HER2 (2+)/ISH+ cohorts. Furthermore, we undertook a comparative assessment of PCR rates and DFS across these 2 cohorts under various NT regimens, both with and without the inclusion of targeted therapy. Our aim was to probe the effectiveness of targeted therapy within these specific cohorts. Additionally, we explored the molecular profiles of the aforementioned subgroups using PAM50 analysis, which allowed us to achieve an augmented understanding of the biology and heterogeneity between HER2 (3+) and HER2 (2+)/ISH+ BC. The findings derived from this study are poised to offer valuable insights for optimizing treatment strategies. Specifically, our results will aid in the identification of patients who are likely to respond favorably to HER2-targeted therapies, thereby enhancing overall patient outcomes.

Methods

This work received ethical approval from the Henan Cancer Hospital, the informed consent requirement was waived owing to the use of anonymized information, and it strictly abided by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) criteria for cohort investigations.²⁵

Sample Population

Two distinct cohorts of patients with HER2-positive early BC receiving NT and surgery were recruited for analysis. Among those excluded from the analysis were patients <18 years or >80 years of age.

Cohort 1 (C1) consisted of 2648 patients selected from 2759 patients who received diagnosis between January 1, 2009 and December 31, 2022, from the Shanghai Jiao Tong University BC Data Base (SJTU-BCDB; which is a nonpublic, multicenter database with over 80 000 records of early-stage BC data from 42 hospitals in 13 provinces across China). One hundred and eleven women were excluded due to the following reasons: (1) missing NT data (n = 68), (2) received primary tumor or axillary surgery prior to NT (n = 31), (3) missing estrogen receptor (ER) or progesterone receptor (PR) status (n = 8), or (4) missing postoperative pathology data (n = 4).

Cohort 2 (C2) consisted of 135 patients with HER2-positive, with associated genomic information acquired from the Henan Cancer Hospital between January 1, 2021 and December 31, 2022. Similar to C1, the C2 patients also received NT, however, their regimen included trastuzumab and pertuzumab.

The C1 and C2 patients were further categorized into 2 subcohorts prior to analysis: (1) patients with IHC HER2 (3+) early BC, (2) patients with IHC HER2 (2+) but ISH amplified early BC. Data analysis was completed between January 1, 2023 and June 30, 2023.

Baseline Demographics

The following baseline demographics were collected and analyzed: age at diagnosis, tumor stage, lymph node status, ER status, PR status, and KI67 status. Also analyzed were 3 independent NAC regimens. Patient age was separated into 2 distinct groups: ≤50 years old and >50 years old. Tumor stage was identified as T1, T2, or T3/4, based on the seventh edition of the American Joint Committee on Cancer criteria. The lymph node status was determined to be N0 (negative imaging or suspicious imaging with negative biopsy), or N+, based on biopsy reports. The hormone receptor (HoR) status was separated into 2 groups: HoR-positive (ER and/or PR positive) and HoR-negative (ER and PR negative). In C1, NT was divided into 3 groups: NAC, NAC plus trastuzumab (NACT), and NAC plus trastuzumab and pertuzumab (NACTP).

Clinical Outcome

The primary endpoint was PCR rate, which represented the lack of residual invasive carcinoma in both the breast and axillary lymph nodes, as evidenced by the final pathological report.

DFS and intrinsic subtyping differences were also compared as secondary endpoints. DFS represented the duration from surgery to local-regional or distant recurrence, a second primary tumor, or death from any cause, whichever occurred first. PAM50 classification via whole RNA sequencing was used for HER2-positive BC subtyping. Lastly, HER2-positive BC was separated into the HER2E and non-HER2E subtypes (which included luminal A, luminal B, normal-like, and basal-like subtypes).

Additionally, we conducted exploratory analysis whereby the gene mutation profiles and enriched signaling networks were compared between patients with HER2 (3+) and HER2 (2+)/ISH+ BC in C2, using a large panel of 1021 target genes.²⁶

Statistical Analysis

All data analyses were performed with SPSS 25.0 or R (R version 4.0.5; R: The R-Project for Statistical Computing, Vienna, Austria). The PCR rate and intrinsic typing differences were assessed via χ² tests. The Kaplan-Meier method was used to generate survival curves, and analysis was done via the log-rank test.

Logistic regression and Cox proportional hazards models were used for odds ratio (OR), hazard ratio (HR), and 95% CI estimations. Finally, 2-sided P < .05 values were set as the significance threshold.

Results

Baseline Demographics

C1 included 2648 women, with a median age of 50 years (IQR, 43-57 years). Among them, 2217 (83.7%) had a HER2 (3+) BC diagnosis, whereas, 431 (16.3%) had a HER2 (2+)/ISH+ BC diagnosis. One thousand and one hundred seventy-two patients (44.3%) were HoR-negative, and 420 patients (15.9%) exhibited reduced KI67 profile (≤20%). One thousand and nine hundred and ninety patients (75.2%) presented with T1 or T2 disease, while 685 patients (25.9%) exhibited node-negative disease. Six hundred and thirty-eight patients (24.1%) underwent NAC without trastuzumab and pertuzumab, whereas, 1231 patients (46.5%) received NAC with trastuzumab (Supplementary Table S1). Table 1 summarizes the baseline demographics of both aforementioned cohorts.

Table 1.

Baseline patient and tumor characteristics among cohort one and cohort 2.

Characteristics	Cohort 1 (n = 2648)			Cohort 2 (n = 135)
	HER2 (3+)	HER2 (2+)/ISH+	P value	HER2 (3+)	HER2 (2+)/ISH+	P value
Age (years)
≦50	1122 (50.6%)	216 (50.1%)	.851	41 (46.1%)	24 (52.2%)	.623
>50	1095 (49.4%)	215 (49.9%)		48 (53.9%)	22 (47.8%)
HoR
Positive	1149 (51.8%)	327 (75.9%)	<.001	44 (49.4%)	31 (67.4%)	.071
Negative	1068 (48.2%)	104 (24.1%)		45 (50.6%)	15 (32.6%)
cT
T1	270 (12.2%)	58 (13.5%)	<.001	3 (3.4%)	2 (4.3%)	.73
T2	1364 (61.5%)	298 (69.1%)		71 (79.8%)	39 (84.8%)
T3/4	583 (26.3%)	75 (17.4%)		15 (16.9%)	5 (10.8%)
cN
N0	584 (26.3%)	101 (23.4%)	.45	24 (27.0%)	6 (13.0%)	.061
N1	1092 (49.3%)	220 (51.0%)		35 (39.3%)	29 (63.0%)
N2/3	541 (24.4%)	110 (25.5%)		30 (33.7%)	11 (23.9%)
KI67
≦20%	331 (14.9%)	89 (20.6%)	.007	2 (2.2%)	7 (15.2%)	.007
>20%	1866 (84.2%)	336 (78.0%)		87 (97.8%)	39 (84.8%)
Unknown	20 (0.9%)	6 (1.4%)		0 (0%)	0 (0%)
Regimen
NAC	518 (23.4%)	120 (27.8%)	<.001	0 (0%)	0 (0%)	NA
NACT	1087 (49.0%)	144 (33.4%)		0 (0%)	0 (0%)
NACTP	612 (27.6%)	167 (38.7%)		89 (100%)	46 (100%)

Abbreviations: HoR, hormone receptor; cT, clinical tumor stage; cN, clinical node status; NAC, neoadjuvant chemotherapy; NACT, neoadjuvant chemotherapy plus trastuzumab; NACTP, neoadjuvant chemotherapy plus trastuzumab and pertuzumab.

C2 included 135 women, with a median age of 49 years (IQR, 42-58 years). Among them, 89 (65.9%) received HER2 (3+) BC diagnosis, 60 patients (44.4%) were identified -negative, 9 patients (6.7%) displayed reduced KI67 profile (≤20%), 115 patients (85.2%) presented with T1 or T2 disease, and 30 patients (22.2%) exhibited node-negative disease. All C2 patients received NAC plus trastuzumab and pertuzumab. Table 1 summarizes the baseline demographics of both aforementioned cohorts.

Pathological Complete Response

In all, 941 (35.5%) patients from C1 achieved PCR. Substantially more patients with HER2 (3+) BC obtained PCR, relative to HER2 (2+)/ISH+ BC [863 (38.9%) of 2217 vs 78 (18.1%) of 431, P < .001] (Fig. 1A), and the difference remained significant following adjustment by clinicopathological factors and NT regimens (OR, 3.16; 95% CI, 2.39-4.17; P < .001; Table 2).

Figure 1.

Pathological complete response rates between HER2(3+) and HER2(2+)/ISH+ groups before(A) and after(B) propensity score matching.

Table 2.

Multivariate analysis for pathological complete response (including HER2 status).

Characteristics	N (PCR%)	Multivariate analysis
		OR	95% CI	P value
Total	2648 (35.5%)
HER2 Status
HER2 (2+)/ISH+	431 (18.1%)	1	Reference
HER2 (3+)	2217 (38.9%)	3.158	2.389-4.173	<.001
Age (years)
≦50	1335 (35.5%)	1	Reference
>50	1313 (35.6%)	0.956	0.805-1.136	.61
HoR
Positive	433 (29.3%)	1	Reference
Negative	508 (43.3%)	1.779	1.496-2.117	<.001
cT
T1	328 (41.5%)	1	Reference
T2	1662 (36.6%)	0.694	0534-0.902	.006
T3/4	658 (29.8%)	0.504	0.373-0.681	<.001
cN
N0	685 (38.2%)	1	Reference
N1	1312 (35.1%)	0.891	0.725-1.095	.271
N2/3	651 (33.5%)	0.824	0.646-1.052	.121
KI67
≦20%	420 (32.4%)	1	Reference
>20%	2202 (36.3%)	1.021	0804-1.297	.866
Unknown	26 (23.1%)	0.694	0.534-0.902	.484
Regimen
NAC	638 (16.5%)	1	Reference
NACT	1231 (34.7%)	2.61	2.041-3.338	<.001
NACTP	779 (52.5%)	6.306	4.848-8.202	<.001

Abbreviations: PCR, pathological complete response; OR, odds ratio; HoR, hormone receptor; cT, clinical tumor stage; cN, clinical node status; NAC, neoadjuvant chemotherapy; NACT, neoadjuvant chemotherapy plus trastuzumab; NACTP, neoadjuvant chemotherapy plus trastuzumab and pertuzumab.

Furthermore, we conducted subgroup analyses to confirm the improved PCR rates in patients with HER2 (3+) BC compared to those with HER2 (2+)/ISH+ BC across most subgroups (Fig. 2). Notably, the impact of HER2 (3+) versus HER2 (2+)/ISH+ status on PCR outcomes exhibited variations based on status and NT regimens, with interaction P values of .027 and <.001, respectively (Fig. 2).

Figure 2.

Subgroup analysis of pathological complete response in patients with HER2(3+) breast cancer vs HER2(2+)/ISH+ breast cancer.

To minimize possible selection bias and imbalance in clinical demographics, propensity score matching analysis was conducted with C1 data to further validate our results. Following a 2:1 propensity score matching for patient age, HoR status, T stage, node status, KI67 status, and NT regimen, we identified 861 patients with HER2 (3+) BC and 431 patients with HER2 (2+)/ISH+ BC for analysis (Supplementary Table S2). The corresponding results corroborated well with the multivariable logistic regression analysis results. In short, relative to patients with HER2 (2+)/ISH+ BC, patients with HER2 (3+) BC still displayed markedly enhanced PCR rate (337 [39.1%] of 861 vs 78 [18.1%] of 431, P < .001; Fig. 1B).

Approximately 16.5%, 34.7%, and 52.5% PCR rates were achieved among NAC-, NACT-, and NACTP-receiving patients, respectively.

Trastuzumab and pertuzumab introduction to NAC augmented PCR rates among patients with HER2 (3+) BC (NACT vs NAC: OR, 2.88; 95% CI, 2.22-3.74; P < .001; NACTP vs NAC: OR, 7.57; 95% CI, 5.72-10.03; P < .001), but not in patients with HER2 (2+)/ISH+ BC (NACT vs NAC: OR, 1.21; 95% CI, 0.62-2.38; P = .58; NACTP vs NAC: OR, 1.72; 95% CI, 0.92-3.24; P = .09; Fig. 3; Table 3).

Figure 3.

Comparision of pathological complete response rates for different neoadjuvant regimens in HER2(3+) and HER2(2+)/ISH+ groups.

Table 3.

Multivariate analysis for pathological complete response according to different HER2 status.

Characteristics	HER2 (3+)			HER2 (2+)/ISH+
	N (PCR%)	OR (95% CI)	P value	N (PCR%)	OR (95% CI)	P value
Age (mean)
≦50	1120 (38.7%)	1 [Reference]		215 (19.1%)	1 [Reference]
>50	1097 (39.2%)	1.023 (0.862-1.213)	.795	216 (17.1%)	0.877 (0.537-1.433)	.601
HoR
Positive	1149 (33.8%)	1 [Reference]		327 (13.8%)	1 [Reference]
Negative	1068 (44.5%)	1.571 (1.323-1.866)	<.001	104 (31.7%)	2.913 (1.733-4.895)	<.001
cT
T1	270 (45.2%)	1 [Reference]		58 (24.1%)	1 [Reference]
T2	1364 (40.5%)	0.827 (0.636-1.076)	.157	298 (18.8%)	0.727 (0.373-1.418)	.35
T3/4	583 (32.2%)	0.577 (0.429-0.776)	<.001	75 (10.7%)	0.375 (0145-0.969)	.043
cN
N0	584 (41.3%)	1 [Reference]		101 (20.8%)	1 [Reference]
N1	1092 (38.8%)	0.903 (0.736-1.109)	.331	220 (16.8%)	0.770 (0.424-1.398)	.391
N2/3	541 (36.6%)	0.822 (0.646-1.045)	.109	110 (18.2%)	0.847 (0.428-1.375)	.632
KI67
≦20%	331 (36.9%)	1 [Reference]		8 9 (15.7%)	1 [Reference]
>20%	1866 (39.4%)	1.116 (0.876-1.421)	.511	336 (18.8%)	1.116 (0.876-1.421)	.375
Unknown	20 (25.0%)	1.071 (0.116-9.880)	.951	6 (16.7%)	0.571 (0.203-1.610)	.289
Regimen
NAC	518 (17.0%)	1 [Reference]		120 (14.2%)	1 [Reference]
NACT	1087 (37.1%)	2.879 (2.219-3.735)	<.001	144 (16.7%)	1.212 (0.617-2.379)	.577
NACTP	612 (60.8%)	7.574 (5.719-10.030)	<.001	167 (22.2%)	1.724 (0.919-3.237)	.09

Abbreviations: HoR, hormone receptor; cT, clinical tumor stage; cN, clinical node status; NAC, neoadjuvant chemotherapy; NACT, neoadjuvant chemotherapy plus trastuzumab; NACTP, neoadjuvant chemotherapy plus trastuzumab and pertuzumab.

Disease-Free Survival

Among the C1 population, following a median follow-up of 29 months (IQR, 17-59), 258 out of 2217 patients (11.6%) from the HER2 (3+) cohort, and 38 out of 431 patients (12.8%) from the HER2 (2+)/ISH+ cohort exhibited local/regional or distant recurrence or a second primary tumor, or expired from any cause.

The 3-year DFS rates were 89.67% among patients with HER2 (3+) BC and 91.26% among patients with HER2 (2+)/ISH+ BC. Moreover, the patients with HER2 (3+) BC experienced comparable DFS relative to the patients with HER2 (2+)/ISH+ BC, both in univariate log-rank (P = .46; Fig. 4A) and multivariate analyses (HR, 1.111; 95% CI, 0.782-1.579; P = .556; Supplementary Table S3).

Figure 4.

Disease-free survival between HER2(3+) and HER2(2+)/ISH+ groups(A), different neoadjuvant regimens in HER2(3+) group(B) and HER2(2+)/ISH+ group(C).

Trastuzumab and pertuzumab introduction to NAC strongly prolonged DFS among patients with HER2 (3+) BC (NACT vs NAC: HR, 0.634; 95% CI, 0.483-0.832; P = .001; NACTP vs NAC: HR, 0.414; 95% CI, 0.229-0.752; P = .004; Fig. 4B), but not among patients with HER2 (2+)/ISH+ BC (NACT vs NAC: HR, 0.899; 95% CI, 0.423-1.866; P = .756; NACTP vs NAC: HR, 1.019; 95% CI, 0.295-3.521; P = .977; Fig. 4C; Table 4).

Table 4.

Multivariate analysis for disease-free survival according to different HER2 status.

Characteristics	HER2 (3+)			HER (2+)/ISH (+)
	HR	95% CI	P value	HR	95% CI	P value
Age (years)
≦50	1	Reference		1	Reference
>50	0.925	0.711-1.204	.75	0.909	0.455-1.816	.787
HoR
Positive	1	Reference		1	Reference
Negative	0.958	0.738-1.245	.75	1.182	0.522-2.676	.689
cT
T1	1	Reference		1	Reference
T2	1.493	0.909-2.451	.113	2.366	0.552-10.142	.246
T3/4	2.073	1.240-3.467	.005	3.59	0.767-16.805	.105
cN
N0	1	Reference		1	Reference
N1	1.375	0.939-2.013	.102	1.665	0.473-5.861	.427
N2/3	2.354	1.588-3.491	<.001	4.469	1.292-15.459	.018
KI67
≦20%	1	Reference		1	Reference
>20%	1.086	0.739-1.597	.674	0.653	0.300-1.422	.283
Regimen
NAC	1	Reference		1	Reference
NACT	0.634	0.483-0.832	.001	0.899	0.423-1.866	.756
NACTP	0.414	0.229-0.752	.004	1.019	0.295-3.521	.977

Abbreviations: HoR, hormone receptor; cT, clinical tumor stage; cN, clinical node status; NAC, neoadjuvant chemotherapy; NACT, neoadjuvant chemotherapy plus trastuzumab; NACTP, neoadjuvant chemotherapy plus trastuzumab and pertuzumab.

Intrinsic Subtype Heterogeneity

Among the C2 population, we used PAM50 analysis to identify 87/89 (97.75%) and 24/46 (52.17%) HER2E subtype among the HER2 (3+) and HER2 (2+)/ISH+ BC cohorts, respectively (P < .001; Supplementary Fig. S1A). In addition, Supplementary Table S4 presents the baseline demographics of the various intrinsic subtypes discussed in this paper. Based on our analysis, the patients with HER2E subtype BC were more prone to being HR-negative, with enhanced PCR rates, relative to the non-HER2E subtype (61.3% vs 16.7%; OR, 6.129; 95%CI, 1.771-21.842; P = .004; Supplementary Fig. S1B and Table S5).

Specific Somatic Mutational Landscape and Mutated Networks Across the HER2 Subgroup

We used a large 1021-gene panel (Supplemental Data File) and targeted BC sequencing to evaluate the genomic information of patients with BC, and the results are summarized in Supplementary Fig. S2. Apart from Her2 mutation, we also identified mutations in the tumor protein 53 (TP53), cyclin-dependent kinase 12 (CDK12), and phosphatidylinositol-4,5-bi-sphosphate 3-kinase catalytic subunit alpha (PIK3CA) genes among 50%, 33%, and 33% of the BC population, and the aforementioned 3 were the most prevalent gene mutations.

Furthermore, HER2 (3+) and HER2 (2+)/ISH+ BCs exhibited distinctive DNA mutation profiles. Aside from HER2 gene amplifications, HER2 (3+) breast tumors displayed augmented CDK12 copy number amplifications (P < .001), and diminished frequent mutations in bromodomain containing-2 (BRD2; P < .01) and epidermal growth factor receptor (EGFR; P < .05; Supplementary Fig. S3A).

In addition, we conducted an in-depth exploration of the common mutational patterns in signaling pathways for each HER2 subgroup. Aside from the ErbB/HER signaling axis, HER2 (3+) breast tumors exhibited a higher frequency of mutations in genes associated with cyclin-CDK signaling (P < .001), Janus kinase-signal transducer and activators of transcription signaling (P < .05), and mitogen-activated protein kinase activity signaling networks (P < .01). Conversely, there was a reduced prevalence of mutations in genes related to vasculogenesis (P < .05) and the canonical WNT signaling pathway (P < .05) in HER2 (3+) tumors when compared to HER2 (2+)/ISH+ tumors (Supplementary Fig. S3B).

Discussion

In this study, we conducted an analysis to assess differences in the response to NT, long-term patient outcomes, and intrinsic subtype heterogeneity between patients with HER2 (3+) and HER2 (2+)/ISH+ early BC. Our findings indicate that HER2 (2+)/ISH+ BC shows lower responsiveness to trastuzumab- and pertuzumab-containing NT compared to HER2 (3+) BC. The effectiveness of neoadjuvant trastuzumab and pertuzumab therapy in HER2 (2+)/ISH+ BC is limited, which can be attributed partially to the intrinsic heterogeneity within these subtypes.

According to the criteria established by the ASCO/CAP,^1-3 both HER2 (3+) and HER2 (2+)/ISH+ BCs are classified as HER2-positive BC. In most clinical studies and treatment guidelines, these 2 cohorts are not clearly distinguished. A previous retrospective investigation with a small sample size of 192 patients showed that among patients with early HER2-positive BC who underwent NAC with trastuzumab and pertuzumab, the PCR rate was 55.1% for patients with HER2 (3+), substantially higher than the 17.6% rate in patients with HER2 (2+)/ISH+ (P < .001).²⁴ In this study, we examined a large number of HER2-positive BC cases, providing more robust statistical power and confirming similar conclusions. Importantly, this study not only included cases that received NAC with trastuzumab and pertuzumab but also cases from an earlier time period who either received NAC alone or NAC with trastuzumab. This allowed us to compare the efficacy of NAC, NAC with trastuzumab, and NAC with trastuzumab and pertuzumab in HER2 (3+) and HER2 (2+)/ISH+ positive populations separately. In the HER2 (3+) population, the addition of trastuzumab or trastuzumab with pertuzumab to NAC resulted in significant improvements in both PCR and DFS, consistent with findings from the NOAH¹⁷ and Neosphere¹⁸ studies. However, in the HER2 (2+)/ISH+ population, we found for the first time that the addition of targeted therapy did not increase PCR rates or improve DFS. There are also ongoing clinical trials (ATTEMPT 2.0 and ELPIS trials) that focus specifically on patients with immunohistochemistry-detected HER2 (3+) as HER2-positive cases.

In the survival analysis, we found no significant difference in DFS between patients with HER2 (3+) and HER2 (2+)/ISH+ BC. This observation could be attributed to a higher proportion of luminal A or luminal B intrinsic subtypes among patients with HER2 (2+)/ISH+. Although intrinsic luminal A and luminal B subtypes of BC are known to be less responsive to NAC and targeted therapy, exhibiting lower PCR rates, they generally have better prognosis and are more likely to benefit from subsequent endocrine therapy.²⁰

Thus far, there has been extensive research on the sensitivity of HER2-targeted therapy, and tumor intrinsic subtypes heterogeneity is considered among the major reasons behind differences in responses to targeted therapy.²⁷ Transcriptomics PAM50-based stratification can effectively delineate between a subset of HER2-positive BC, namely the HER2E subtype, which is highly responsive to targeted therapy. The PAM50-defined HER2E intrinsic subtype is also strongly correlated with enhanced HER2 transcript levels, as well as augmented PCR rate, relative to other subtypes.^22,28,29 Herein, we demonstrated an intricate relationship between the HER2 protein expression status and intrinsic subtypes. Among patients with HER2 (3+), 97.75% harbored the HER2E subtype, which was considerably higher than the 52.17% observed among patients with HER2 (2+)/ISH+. This likely explained why patients with HER2 (3+) were highly responsive to targeted therapy. Moreover, this finding suggested that, in clinical practice, the integrated examination of the HER2 protein expression status and PAM50-based intrinsic subtyping can greatly assist the selection of patient population that may truly benefit from targeted therapy. In this report, somatic mutational profiling of a 1021-gene panel NGS was used to illustrate that HER2 (3+) BC possessed more CDK12 amplifications. CDK12 is coded by chromosome 17q12, which is a close neighbor of the HER2 gene. Hence, CDK12 is frequently amplified with HER2.^30,31 CDK12 is responsible for the alternative splicing of genes related to the DNA damage response axis as well as BC pathogenesis.³² Several investigations validated the close relationship between CDK12 and targeted therapy resistance in patients with HER2+ BC.^33,34 Moreover, our findings generated some interest into the significance of CDK12 among patients with HER2 (3+) and HER2 (2+)/ISH+ BC, as well as its relationship to sensitivity to targeted treatment.

Limitations

This work has several limitations. First, it is retrospective, thus, there may be some unintentional selection bias and confounding factors. We recommend additional prospective investigations, involving large sample sizes, to validate our results. Second, the patient follow-up time was relatively short, which may have limited our ability to accurately evaluate long-term patient outcomes. Thus, we warrant future investigations, involving longer follow-up durations to better elucidate the NAC influence on patient OS. Thirdly, in the context of neoadjuvant research, event-free survival (EFS) may be a better survival endpoint than DFS, but due to the retrospective nature of the data, information on the efficacy assessment during neoadjuvant treatment is missing, making it impossible to calculate accurate EFS data. However, DFS is also a commonly used survival metric in neoadjuvant research. Lastly, this study examined patients from certain institutions, which may have inadvertently limited the generalizability of our findings to other populations.

Conclusions

This cohort investigation provides compelling evidence of a significant difference in the response to NT between patients with HER2 (3+) and HER2 (2+)/ISH+. Patients with HER2 (2+)/ISH+ encounter challenges in deriving benefit from neoadjuvant targeted therapy, potentially due to intrinsic subtype differences. Based on our findings, we recommend considering HER2 protein expression when making neoadjuvant targeted treatment recommendations for patients with early-stage HER2-positive BC. Reevaluation of the value of targeted therapy in patients with HER2 (2+)/ISH+ is necessary. Additionally, further investigation is warranted to explore the potential value of ADC drugs in the treatment of patients with HER2 (2+)/ISH+, given their effectiveness in HER2-positive and low HER2 expression cases.

Funding

This study was funded by the Training Program for Young and Middle-aged Health and Technology Innovation Leaders in Henan Province (YXKC2022005,LJRC2023003),the National Natural Science Foundation of China(No. 82372712),and the Science and Technology Research Projects of Henan province (No. 232102311125).

Conflict of Interest

The authors of the article declare that there are no conflicts of interest related to this manuscript.

Author Contributions

Conception/design: D.J., G.L., H.D., Z.L. Provision of study material or patients: J.W., J.Z., Y.H. Collection and/or assembly of data: J.W., J.Z., Y.H. Data analysis and interpretation: Y.Z., X.G., Z.L. Manuscript writing: D.J., G.L., H.D., Y.H. Final approval of manuscript: All authors.

Data Availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Ethics Approval and Consent to Participate

This study was approved by the ethics committees of Henan Cancer Hospital and was conducted in accordance with the Declaration of Helsinki.