An exploration of the optimal combination chemotherapy regimen based on neoadjuvant therapy containing pyrotinib for HER2-positive breast cancer: A multicenter real-world study

Highlights

• •Based on neoadjuvant therapy containing pyrotinib, an anthracycline-free regimen can yield pCR rates no less favorable than those of anthracycline-containing regimens.
• •Platinum-containing, long-cycle taxane regimens appear to achieve superior efficacy under anthracycline-removed conditions.
• •Estrogen receptor (ER) status and HER2 status were independent predictors of tpCR.

Abstract

Background

The combination of pyrotinib (Py) with cytotoxic agents proved to be effective in early human epidermal growth factor receptor 2 (HER2)-positive breast cancer (BC). However, the optimal chemotherapy regimen is unknown. This study attempts to explore it from real-world research data.

Methods

Information was collected from patients with early-stage HER2-positive BC from 23 centers across the country. They were categorized into the anthracycline group (A group) and non-anthracycline group (non-A group). Patients in the non-A group were further categorized into the platinum group and non-platinum group and the short-cycle (≤4 cycles) taxane group and long-cycle (>4 cycles) taxane group. Total pathological complete response (tpCR, ypT0/is ypN0) and breast pathological complete response (bpCR, ypT0/is) rates were assessed.

Results

A total of 107 patients were enrolled. Postoperative pathology indicated a tpCR rate of 36.8 %, a bpCR rate of 42.1 % in the A group, the non-A group had a tpCR rate of 47.8 %, and a bpCR rate of 53.6 %, with P-values of 0.273 and 0.254, respectively. In the long-cycle taxane group, the tpCR and bpCR rates were 60.8 % and 66.7 %, respectively. In the short-cycle taxane group, the tpCR and bpCR rates were 11.1 % and 16.7 %, respectively (both P<0.001). The platinum group had higher tpCR rate (62.9 % vs. 32.4 %, respectively; P = 0.011) and bpCR rate (65.7 % vs. 41.2 %, respectively; P = 0.041).

Conclusion

As for a neoadjuvant therapy regimen with Py, an anthracycline-free regimen is feasible. Besides, platinum-containing, long-cycle taxane regimens appear to achieve superior efficacy under anthracycline-removed conditions.

Introduction

Human epidermal growth factor receptor 2 (HER2)-positive breast cancer (BC) is a highly invasive and recurrent BC, accounting for about 15 %–20 % of all BC [1]. Currently, neoadjuvant therapy is widely used for early high-risk and locally advanced HER2-positive BC. Neoadjuvant therapy not only reduces the extent of the lesion and improves breast and axillary conservation rates, but also provides individualized drug sensitivity information, which guides subsequent treatment. Moreover, studies have shown that patients who achieve pathological complete response (pCR) have a better long-term prognosis [2].

Results from the NOAH phase 3 clinical trial showed that trastuzumab (H) combined with chemotherapy could significantly improve the pCR rate (38.5 % vs. 19.5 %, respectively) and 3-year event-free survival (71 % vs. 56 %, respectively) in patients with HER2-positive BC [3]. Since then, H and other anti-HER2 antibody drugs have been added to the neoadjuvant therapy. Double-target therapy has taken neoadjuvant therapy to the next level in HER2-positive BC. In the NeoSphere phase 2 trial, 4 cycles of docetaxel (T) in combination with an H and pertuzumab (P) regimen achieved a pCR rate of 45.8 %, which was much higher than a rate of 29.0 % with a TH regimen and 24.0 % with a TP regimen [4]. In recent years, orally available small-molecule tyrosine kinase inhibitors (TKIs) have been developed which have the advantages of being orally available, having low cardiotoxicity, and being able to penetrate the blood-brain barrier, as compared with large-molecule targeted drugs [5]. The NeoALTTO study involved a preliminary exploration of the application of TKIs in the neoadjuvant therapy for HER2-positive BC. In this study, patients with HER2-positive BC were randomized to be administered lapatinib (La), H, or La+H with sequential 12-week paclitaxel treatment, and the results indicated that the La+H regimen had better pCR rate than the H regimen (51.3 % vs. 29.5 %, respectively) [6]. Besides, the PHEDRA phase 3 trial compared the efficacy of a pyrotinib (Py)+TH regimen with the TH regimen, and the results showed that the pCR rate for the Py+TH group was significantly higher than that of the control group (41 % vs. 22 %, respectively) [7]. Thus, combination with TKIs can also improve the pCR rate in HER2-positive BC compared with single H therapy.

However, in the context of anti-HER2, especially dual anti-HER2 neoadjuvant therapy, the choice of paired chemotherapeutic agents and whether de-escalated therapy is possible has become a new challenge. Both the NeoSphere trial and the PEONY trial examine a combination of a dual-targeted regimen in combination with docetaxel supplemented with 3 cycles of postoperative FEC (5-F: fluorouracil, E: epirubicin, C: cyclophosphamide) chemotherapy [4,8]. In the TRYPHAENA phase 2 trial, the FEC-THP regimen and TCbHP (Cb: carboplatin) regimen had similar pCR rates, which suggests that an anthracycline-free scheme is feasible [9]. Furthermore, the TRAIN-2 study directly compared the efficacy between the anthracycline group (FEC+THP-TCbHP regimen) and the non-anthracycline group (TCbHP regimen); interestingly, the results showed that there was no difference in the pCR rates between the two groups. However, in terms of toxicities, febrile neutropenia and cardiologic toxicity were significantly higher in the anthracycline group than in the non-anthracycline group [10]. Previous studies have found an increased risk of cardiac dysfunction, especially heart failure, in patients when anthracyclines are combined with H [[11], [12], [13], [14]]. Long-term follow-up results from the TRYPHAENA and TRAIN-2 studies have also shown that anthracycline exemptions do not affect long-term survival [15,16].

Py can irreversibly bind to three receptor sites: HER1, HER2, and HER4 [17]. Several clinical trials have reported that neoadjuvant therapy regimens containing Py can also achieve high pCR rates [7,[18], [19], [20], [21]] (Table 1). However, open questions remain: what is the optimal chemotherapy dosing regimen for Py? Is an anthracycline-free neoadjuvant therapy feasible? If so, what is the optimal chemotherapy regimen? Our study starts with real-world data on the efficacy of Py combined with anthracyclines or combined with an anthracycline-free regimen as neoadjuvant treatment for early or locally advanced HER2-positive BC. On the basis of this, we preliminarily explored the optimal chemotherapy with neoadjuvant therapy regimens containing Py.

Table 1.

Neoadjuvant trials involving pyrotinib.

Number of the study	Phase	Participants	N	Regimens	Definition of pCR	pCR rate (%)
ChiCTR1900022293	Ⅱ	stage Ⅰ∼Ⅲ HER2+BC patients	20	Py+4EC-4TH	ypT0/is, ypN0	73.7
NCT04152057	Ⅱ	stage Ⅱ∼Ⅲ HER2+BC patients	21	Py+4TH	ypT0/is, ypN0	57.1
NCT04126525	Ⅱ	stage Ⅱ∼Ⅲ HER2+BC patients	53	Py+4TPH	ypT0, ypN0	69.81
NCT03735966	Ⅱ	stage Ⅱ∼Ⅲ HER2+BC patients	75	Py+6TCbH	ypT0/is, ypN0	55.1
NCT03588091	Ⅲ	stage Ⅱ∼Ⅲ HER2+BC patients	355	Py+4TH	ypT0/is, ypN0	41.0

BC, breast cancer; Py, pyrotinib; E, epirubicin; C, cyclophosphamide; T, taxanes; H, trastuzumab; P, cisplatin; Cb, carboplatin.

Methods

Study design and patients

This multicenter, real-world study retrospectively analyzed information on female patients with HER2-positive BC administered neoadjuvant Py combined with chemotherapy from November 2018 to June 2023 from 23 centers throughout China. The main inclusion criteria were as follows: 1. histopathologically confirmed clinical stage II-III primary HER2-positive invasive BC (core needle aspiration biopsy or vacuum-assisted breast biopsy); 2. at least one measurable lesion according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [22]; 3. received neoadjuvant therapy containing Py, with at least 1 cycle of Py; 4. received surgery and completed postoperative pathology. Fig. 1 shows the detailed process.

Fig. 1.

Flow diagram of the patient population.

Clinical and pathological evaluation

The baseline status for HER2, estrogen receptor (ER), progesterone receptor (PR), and Ki-67 was assessed using immunohistochemistry (IHC). The HER2 IHC 2+ specimens were tested using fluorescence in situ hybridization (FISH). HER2 IHC 3+ or 2+ along with FISH positive were considered positive for HER2. With respect to the 2010 ASCO/CAP guidelines [23], ER or PR positive was defined as having ≥1 % of tumor cell nuclei immunoreactive; negative was defined as having <1 % of tumor cell nuclei immunoreactive. Ki-67 was assessed as the percentage of positive immunostained tumor cell nuclei. Clinical stage was assessed according to the eighth edition of American Joint Committee on Cancer tumor-node-metastasis (TNM) staging system [24].

Treatment

The standard dose of oral Py is 400 mg/d. The starting dose of the neoadjuvant combination regimen and Py is determined by the local investigator. Dose reduction, interruption or discontinuation of Py based on adverse events is permitted.

Data collection

Relevant information was extracted from medical records, including demographics, baseline characteristics, neoadjuvant therapy, pathological findings, and surgical information. Pathology reports are completed by local pathologists.

Outcome

Neoadjuvant efficacy was evaluated according to RECIST version 1.1 [22] Patients underwent surgery at the end of the neoadjuvant course. The primary endpoint was the total pathological complete response (tpCR) rate, defined as the proportion of patients with no residual invasive tumor cells in breast and axillary lymph nodes (ypT0/is ypN0). The secondary study endpoint was the breast pathological complete response (bpCR) rate, defined as the proportion of patients without any residual invasive cancer in the resected breast specimen (ypT0/is).

Statistical analysis

Comparisons of baseline characteristics between subgroups were performed by the two independent samples test for continuous variables, the two independent samples Wilcoxon signed rank sum test for ranked variables, and the Pearson χ2 test or Fisher exact probability test for categorical variables. Logistic regression was used for correlation factor analysis. Pearson χ2 test and Cochran-Mantel-Haenszel (CMH) test were used to assess the pCR rates in the anthracycline group (A group) and the non-anthracycline group (non-A group), and the CMH test adjusted for the effects of the stratification variables ER status and HER2 status. Clopper Pearson exact test was used to calculate 95 % confidence intervals. P < 0.05 was considered statistically significant and all tests were two-sided. All data analysis was performed using IBM SPSS 26.0.

Results

Baseline characteristics

A total of 107 patients were finally enrolled in our study (Fig. 1). The neoadjuvant therapy regimens can be found in Table 2. Of the 107 patients, 4 (3.7 %) completed only 1 cycle of Py, 6 (5.6 %) completed 2 cycles of Py, 2 (1.9 %) completed 3 cycles of Py, 22 (20.6 %) completed 4 cycles of Py, 10 (9.3 %) completed 5 cycles of Py, 48 (44.9 %) completed 6 cycles of Py, 2 (1.9 %) completed 7 cycles of Py, and 13 (12.1 %) completed 8 cycles of Py. Further, 38 patients had a neoadjuvant therapy regimen containing anthracyclines, and 69 patients were not administered anthracyclines. Differences in baseline clinical characteristics between the two groups are shown in Table 3. There was a greater proportion of postmenopausal patients in the A group than the non-A group (52.6 % vs. 42.0 %, respectively). 26 (68.4 %) patients in the A group were ER positive, but there were more patients who were ER negative in the non-A group (n = 44, 63.8 %). More than 80 % of patients in both the A group and the non-A group were classified as HER2 IHC 3+ (84.2 % vs. 84.1 %, respectively).

Table 2.

Neoadjuvant therapy regimens.

Regimens	Number of people (%)	Median treatment cycles (range)
Dual-target
TCbH+Py or TLH+Py	32 (30.0 %)	6 (2–6)
TH+Py	27 (25.2 %)	6 (1–7)
AC-TH+Py	17 (15.9 %)	8 (6–8)
T-ACH+Py	13 (12.1 %)	8 (8–8)
ACH+Py	3 (2.8 %)	4 (4–4)
ATH+Py	1 (0.9 %)	1
TCH+Py	1 (0.9 %)	4
AC-TCbH+Py	1 (0.9 %)	7
AI+H+Py	1 (0.9 %)	6
OFS+AI+H+Py	1 (0.9 %)	2
N+Pb+H+Py	1 (0.9 %)	2
AC+Py, then switched to TCbHPa	1 (0.9 %)	8
Single-target or triple-target
X+Py	1 (0.9 %)	6
TAC+Py	1 (0.9 %)	8
TC+Py	1 (0.9 %)	6
TCb+Py	1 (0.9 %)	6
T+X+Py	1 (0.9 %)	6
OFS+AI+Py	1 (0.9 %)	6
TCbHP+Py	1 (0.9 %)	6
AC-THP+Py	1 (0.9 %)	8

T, taxane (docetaxel, albumin-bound paclitaxel, or paclitaxel); Cb, carboplatin; H, trastuzumab; L, Lobaplatin; Py, pyrotinib; A, anthracycline (epirubicin, pirarubicin, or doxorubicin); C, cyclophosphamide; AI, aromatase inhibitor (anastrozle or exemestane); OFS, ovarian function suppression (goserelin); N, vinorelbine; Pb, cisplatin; X, capecitabine; P, pertuzumab.

^aThis patient received 2 cycles of AC+Py regimen, then switched to 6 cycles of TCbHP regimen for unknown reason.

Table 3.

Baseline characteristics.

	Non-anthracycline group (n=69)	Anthracycline group (n=38)	P value
Age (Average)	50.17	50.66	0.692
Menopausal status			0.292
Premenopausal	40 (58.0 %)	18 (47.4 %)
Postmenopausal	29 (42.0 %)	20(52.6 %)
Clinical tumor stage			0.337
T1	7 (10.2 %)	2 (5.2 %)
T2	41 (59.4 %)	22 (57.9 %)
T3	15 (21.7 %)	9 (23.7 %)
T4	6 (8.7 %)	5 (13.2 %)
Clinical nodal stage			0.338
N0	9 (13.1 %)	5 (13.2 %)
N1	35 (50.7 %)	23 (60.5 %)
N2	11 (15.9 %)	6 (15.8 %)
N3	14 (20.3 %)	4 (10.5 %)
Clinical stage			0.721
Ⅱ	37 (53.6 %)	19 (50.0 %)
Ⅲ	32 (46.4 %)	19 (50.0 %)
ER status			0.001
Negative	44 (63.8 %)	12 (31.6 %)
Positive	25 (36.2 %)	26 (68.4 %)
PR status			0.178
Negative	42 (60.9 %)	18 (47.4 %)
Positive	27 (39.1 %)	20 (52.6 %)
HER2 status			0.984
IHC 3+	58 (84.1 %)	32 (84.2 %)
IHC 2+ and FISH+	11 (15.9 %)	6 (15.8 %)
Ki-67 %			0.958
<30	13 (18.8 %)	7 (18.4 %)
≥30	56 (81.2 %)	31 (81.6 %)

ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; IHC, immunohistochemistry; FISH, fluorescence in situ hybridization.

Outcomes

After surgery, the pathological results indicated that 53 (49.5 %; 95 % CI 39.7 %-59.4 %) patients achieved bpCR, and 47 (43.9 %; 95 % CI 34.3 %-53.9 %) patients achieved tpCR. In the non-A group, 37 patients achieved bpCR (53.6 %; 95 % CI 41.2 %-65.7 %) and 33 patients achieved tpCR (47.8 %; 95 % CI 35.6 %-60.2 %). In the A group, 16 patients achieved bpCR (42.1 %; 95 % CI 26.3 %-59.2 %), and 14 patients achieved tpCR (36.8 %; 95 % CI 21.8 %-54.0 %). Although the difference between the two groups did not reach statistical significance (P-values were 0.254 and 0.273, respectively), it can be seen that the bpCR and tpCR rates for the non-A group were numerically higher than those of the A group (Fig. 2a).

Fig. 2.

Breast pathological complete response (bpCR) rate and total pathological complete response (tpCR) rate.

(a)The bpCR rate and tpCR rate of the overall population,non-anthracycline group and anthracycline group. (b) Subgroup analysis of tpCR rate with respect to estrogen receptor (ER) status. (c) Subgroup analysis of tpCR rate with respect to the human epidermal growth factor receptor 2 (HER2) status. bpCR, breast pathological complete response; tpCR, total pathological complete response.

Ten factors potentially influencing tpCR including whether using anthracyclines were applied to the univariate and multivariate analyses. Among them, ER status (OR = 0.140; 95 % CI 0.034–0.583; P = 0.007) and HER2 status (OR = 10.318; 95 % CI 1.937–54.967; P = 0.006) were independent influencing factors on tpCR. After adjusting for all factors in the multivariate analysis (age, menopausal status, clinical tumor stage, clinical nodal stage, clinical stage, ER status, PR status, HER2 status, and Ki-67 %), there was still no significant difference in the tpCR rate between the A group and the non-A group (OR = 1.058; 95 % CI 0.383–2.927; p = 0.913) (Table 4).

Table 4.

Factors influencing total pathological complete response (tpCR) rate.

Variables	Univariate analysis		Multivariate analysis
	OR (95 % CI)	P value	OR (95 % CI)	P value
Age	1.127 (0.519–2.445)	0.762	0.775 (0.244–2.463)	0.665
<50
≥50
Menopausal status	1.253 (0.582–2.698)	0.564	1.168 (0.388–3.516)	0.783
Premenopausal
Postmenopausal
Clinical tumor stage	1.114 (0.494–2.509)	0.795	1.336 (0.322–5.547)	0.690
T1–2
T3–4
Clinical nodal stage	0.384 (0.119–1.235)	0.108	0.326 (0.074–1.431)	0.137
N0
N+
Clinical stage	0.941 (0.438–2.021)	0.875	1.022 (0.257–4.066)	0.975
Ⅱ
Ⅲ
ER status	0.221 (0.097–0.506)	<0.001	0.140 (0.034–0.583)	0.007
Negative
Positive
PR status	0.347 (0.155–0.777)	0.010	0.964 (0.264–3.525)	0.956
Negative
Positive
HER2 status	7.500 (1.620–34.712)	0.010	10.318 (1.937–54.967)	0.006
IHC 3+
IHC 2+ and FISH+
Ki-67 %	0.740 (0.279–1.960)	0.545	0.376 (0.110–1.285)	0.119
<30
≥30
Group	0.636 (0.283–1.432)	0.275	1.058 (0.383–2.927)	0.913
Non-anthracycline group
Anthracycline group

ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; IHC, immunohistochemistry; FISH, fluorescence in situ hybridization.

Subgroup analysis

According to the multivariate analysis results, the subgroup analysis focused on ER and HER2 status. Overall, 7 patients with ER negative (58.3 %; 95 % CI 27.7 %-84.4 %) and 7 patients with ER positive (26.9 %; 95 % CI 11.6 %-47.8 %) achieved tpCR in the A group. In the non-A group, 27 patients with ER negative (61.4 %; 95 % CI 45.5 %-75.6 %) and 6 patients who were ER positive (24.0 %; 95 % CI 9.4 %-45.1 %) achieved tpCR. The tpCR rates were higher in ER negative BC, whether administered an anthracycline-containing or anthracycline-free regimen (Fig. 2b). Furthermore, 40.6 % (95 % CI 23.7 %-59.4 %) of patients with HER2 IHC 3+ in the A group achieved tpCR, as compared with 16.7 % (95 % CI 4 %-64.1 %) of patients with HER2 IHC 2+ and FISH amplification. In the non-A group, tpCR rates were 55.2 % (95 % CI 41.5 %-68.3 %) and 9.1 % (95 % CI 2 %-41.3 %) in HER2 IHC 3+ and HER2 IHC 2+, respectively. Among patients treated with anthracycline-containing and anthracycline-free therapies, the tumor response with neoadjuvant therapy was better in HER2 IHC 3+ than in HER2 IHC 2+ (Fig. 2c).

Correlation between chemotherapy regimen and pathologic responses in the non-A group

There was no significant difference of both the tpCR and bpCR rates between the A group and the non-A group (P-value > 0.05), regardless of whether confounders were corrected. However, after omitting anthracyclines, it is not clear which chemotherapy regimen would have a greater benefit for patients. Thus, we further grouped patients in the non-A group according to whether the neoadjuvant regimen contained platinum or how many cycles of taxanes patients used. Finally, 35 patients were in the platinum group, and 34 were in the non-platinum group. The bpCR rate of the platinum group was significantly higher (65.7 %; 95 % CI 47.8 %-80.9 %) than that of the non-platinum group (41.2 %; 95 % CI 24.6 %-59.3 %), with a P-value of 0.041. Similarly, tpCR rates demonstrated a similar trend, 62.9 % (95 % CI 44.9 %-78.5 %) with platinum vs. 32.4 % (95 % CI 17.4 %-50.5 %) without platinum (P = 0.011) (Fig. 3a). To explore whether the administration of a different number of taxane cycles would affect the efficacy of the anthracycline-free regimen, we performed a statistical analysis on the use of taxane in the non-A group (Table 5); and sought a taxane cycle cut-off value of 4.5 based on the receiver operator characteristic curve (ROCs) (Fig. 3b); and categorized patients into a short-cycle (≤4 cycles) taxane group and long-cycle (>4 cycles) taxane group. First, we compared the pCR rates for patients administered taxane and Py for >4 cycles versus those administered both taxane and Py for ≤4 cycles. We observed that patients administered >4 cycles compared with those in the ≤4 cycles group had higher bpCR rate (66.0 % vs. 23.1 %, respectively; P = 0.006) and tpCR rate (59.6 % vs. 15.4 %, respectively; P = 0.005) (Fig. 3c). As only five patients were administered >4 Py cycles but ≤4 taxane cycles, and similarly, only four patients were administered ≤4 cycles of Py but >4 taxane cycles, we included this group of patients in the overall analysis. Taking into account only the number of taxane cycles, patients administered ≤4 taxane cycles had much lower bpCR rates and tpCR rates than patients administered >4 taxane cycles (16.7 % vs. 66.7 %, respectively; P < 0.001; and 11.1 % vs. 60.8 %, respectively; P < 0.001) (Fig. 3d).

Fig. 3.

Breast pathological complete response (bpCR) rate and total pathological complete response (tpCR) rate in the non-anthracycline group.

(a) The bpCR rate and tpCR rate of the non-platinum group and platinum group. (b) Receiver operator characteristic curve (ROCs)of taxane cycles. (c) The bpCR rate and tpCR rate of patients administered both taxane and pyrotinib for ≤ 4 cycles group and for > 4 cycles group. (d) The bpCR rate and tpCR rate of patients administered taxane for ≤ 4 cycles group and for> 4 cycles. bpCR, breast pathological complete response; tpCR, total pathological complete response.

Table 5.

Details regarding taxanes administration in the non-anthracycline group.

Cycles of taxanes administration	Number of people	Percentage (%)
No taxanes used	5	7.2
1	1	1.5
2	2	2.9
3	1	1.5
4	9	13.0
5	5	7.2
6	45	65.2
7	1	1.5

Discussion

Recently, with the ongoing development of anti-HER2 agents, there has been a significant increase in the variety of targeted and chemotherapeutic agents available for combination therapy. Neoadjuvant regimens incorporating Py have demonstrated impressive pathologic complete response rates in phase 2 and 3 clinical trials. Currently, chemotherapy regimens involving Py typically include trastuzumab and other large-molecule targeted drugs. Examples of such regimens are TCbHP, THP, AC-THP, and TCH, among others [26]. As a TKI, the binding sites, mode of administration, and adverse effects differ from those of large-molecule drugs; therefore, compatible chemotherapy regimens should not be adopted wholesale.

Our study directly compared the neoadjuvant efficacy of Py combined with anthracyclines or combined with an anthracycline-free regimen. The results showed no difference in tpCR rates between the non-A group and the A group, which were 47.8 % and 36.8 %, respectively, P = 0.273. Consistent with previous studies, omitting anthracyclines did not affect the pCR rate [10]. The PHEDRA trialshowed that 4 cycles of Py+TH neoadjuvant therapy, as validated by independent centers, resulted in a tpCR rate of 41 % and a bpCR rate of 43.8 % [7]. Besides, the Panphila trial evaluated the neoadjuvant efficacy of 6 cycles of a Py+TCb regimen, and the tpCR rate was 55.1 % [19]. We can see that neither of the two abovementioned studies used anthracyclines, and the pCR rates were still substantial. However, in another phase 2 clinical trial, the 8-cycle regimen (Py+EC-TH) achieved a pCR rate of 73.7 %. Although the pCR rate in this study was the highest to date, the study enrolled only 20 patients (20 enrolled, 19 completed surgery), included a longer duration of medication (8 cycles), and enrolled a greater percentage of hormone receptor (HR)-negative patients (75 %) [21]. Chronic dose-cumulative cardiotoxicity of anthracyclines is a current clinical concern. A meta-analysis showed that anthracycline-free and anthracycline-containing regimens had similar efficacy under the dual-targeting condition of treponemal, but anthracycline-containing regimens had more pronounced toxicities, especially cardiotoxicity [25]. A study examined the Py+docetaxel+doxorubicin liposomal+cyclophosphamide (TAC) regimen demonstrated good efficacy and safety, with no increase in cardiotoxicity [26]; another single-arm clinical trial corroborated this result [27]. However, it is unknown whether this result would change if trastuzumab is added, and we know that the cardiotoxicity of large-molecule targeted drugs is much higher than that of TKIs. In our study, more than 90 % of patients were concurrently administered trastuzumab. From the PHEDRA and other studys, we can also find that the study designers tried to avoid combining anthracyclines. Combining the results of previous studies with ours suggests that an anthracycline-free chemotherapy regimen can achieve good efficacy while having a good safety profile.

Anthracyclines and taxanes are currently the two cornerstones of BC treatment, and after de-anthracycline, taxanes have an unassailable position. As in our study, most patients who did not use anthracyclines were administered taxanes (64/69, 92.8 %). However, is taxane alone sufficient to fulfill the requirement for efficacy? Is there a need to combine other chemotherapy drugs? The tpCR was 62.9 % for the platinum-containing group and only 32.4 % for the platinum-free group in this study, seeming to infer as such to us. The BCIRG006 study [28] found that an anthracycline plus cyclophosphamide sequential docetaxel plus trastuzumab (AC-TH) regimen was consistent with the disease-free survival (DFS) benefit of a docetaxel plus carboplatin plus trastuzumab (TCbH) regimen, but with a substantial reduction in cardiac adverse events with the TCbH regimen. We can see that TCbH is an effective alternative regimen. Both the KRISTINE prospective study and a retrospective study performed compared the pCR rates of a TCbHP regimen with those of a THP regimen, and the results of both studies demonstrated that the addition of carboplatin to taxanes was effective at improving the pCR rate [29,30]. In some studies, higher pCR rates were obtained with taxane pairings alone. Another study found that a 4-cycle Py+TH regimen achievd a pCR rate of 57.1 %, but the study used taxane weekly [18]. Previous studies have shown that taxane dose-dense chemotherapy regimens are effective at improving the pCR rate of neoadjuvant therapy [31,32]. Similarly, in the ADAPT HER2-positive/HR-negative study, a pCR rate of 90 % could be achieved with 12 weeks of paclitaxel in combination with trastuzumab and patuzumab, but this study was only for HR-nagative patients and used a weekly paclitaxel regimen [33].

Our findings also showed that the long-cycle taxane group had a significantly higher tpCR rate than the short-cycle taxane group. Currently, most studies investigated 4- and 6-cycle regimens in combination with taxanes and target drugs [4,8,29]. In the KRISTINE study and TRYPHAENA studies, a 6-cycle TCbHP regimen attained a pCR rate of 55.7 % and 51.9 %, respectively [9,34] . In the TRAIN-2 clinical trial, a 9-cycle TCbHP regimen resulted in a pCR rate of 68 %, surpassing the 60 % mark [10]. In the TBCRC023 study, patients administered La in combination with trastuzumab had a higher pCR rate in the 24-cycle group than the 12-cycle group (28 % vs. 12 %, respectively) [35]. Above all, longer regimens appear to be more conducive to achieving pCR. Although the 4-cycle THP regimen is also included in the Chinese Society of Clinical Oncology (CSCO) guidelines for BC diagnosis and treatment, the level of recommendation is still lower than the 6-cycle THP regimen. Currently, the 4-cycle THP regimen is more commonly used as a clinical study design in addition to being a treatment option for patients who have poor tolerance. In conclusion, a long-cycle taxane-based regimen is more favorable for achieving pCR if tolerated.

Notably, in the total population, both ER negative and HER2 IHC 3+ were independent correlates of higher tpCR rates. In terms of the molecular mechanism, there is a bidirectional crossover between HER2 and ER signaling pathways, and ER activation is an important cause of resistance to HER2-positive BC-targeted therapy [36,37] . It is also evident from trials such as NeoSphere, KRISTINE, TBCRC006, and Panphila that patients with ER-positive/HER2-positive BC seem to benefit less from targeted therapy [4,19,29,38] . Patients with IHC 3+ or IHC 2+ and FISH amplification are considered HER2-positive for targeted therapy in current clinical guidelines [39]. However, the pCR rate in the Panphila study was 60.7 % in IHC 3+ and only 30.8 % in IHC2+ [19]. In addition, Trastuzumab did not improve DFS in patients with IHC-/FISH-amplified administered trastuzumab in the N9831 study, whereas there was no difference in DFS between IHC 3+/FISH- and IHC 3+/FISH+, suggesting that overexpression of HER2 protein is critical for anti-HER2 therapy [40]. Furthermore, retrospective studies have also shown that the level of HER2 protein expression correlates with pCR in targeted therapies containing trastuzumab, with patients who were IHC 3+ being more inclined to obtain pCR [41]. More than 80 % of the patients we included in our analysis were HER2 IHC 3+, suggesting that we prefer to use the Py combined with H regimen for patients who are HER2 IHC 3+ in the real world.

To the best of our understanding, this represents the inaugural endeavor aimed at discovering the most effective chemotherapy regimen combination featuring Py. When discussing the optimal neoadjuvant chemotherapy protocols for HER2-positive BC, several studies have primarily focused on contrasting the various types and combinations of anti-HER2 antibody-based medications [[42], [43], [44], [45], [46]]. Usually, neoadjuvant chemotherapy regimens in theses studies were H+chemotherapy(CT), HP+CT, trastuzumab emtansine (T-DM1)+CT, TKIs+CT, TKI+H+CT, regimens contained Py were seldom enrolled. And cytotoxic drugs were not compared by categaery in these studies. In addition, there are also investigations that have delved into determining the most suitable treatment cycles for specific neoadjuvant regimens or evaluating the efficacy of different types of a certain cytotoxic drug [[47], [48], [49]]. However, neoadjuvant chemotherapy regimens were fixed, the varity of combinations of durgs was not explored. HER2-positive BC is highly heterogeneous, not all patients are suitable for HER2-targeted monoclonal antibodies or antibody-drug conjugates (ADCs), TKIs like Py provides more choices and possible for HER2-positive BC. Currently, the combination of Py with chemotherapy drugs is various and has no standard. Proceed from the reality, our study analyzed the effect of different chemotherapy drugs on the regimens contained Py in detail, especially the anthracyclines. Furthermore, after omitting anthracyclines, we explored an effective alternative considering both the cytotoxic drugs and the treatment cycles. In this sense, our study provides a clue for the design of clinical trials in the future, and also a reference for the selection of chemotherapy regimens containing Py in the clinic.

Our study also has some limitations. First, this was a retrospective study in which bias is unavoidable. In an effort to minimize potential biases and confounding factors in roder, we performed a multifactorial analysis on the collected data, and corrected for the effect of confounding factors on the results. Second, as there was too much missing information on histopathologic grading of BC, we were unable to include that factor in the analysis. Third, the incomplete documentation of adverse events prevented analysis of the safety of the non-A and A groups that is an important consideration for neoadjuvant efficacy. In addition, long-term survival data and quality of life scores were lacking. Fourth, our study included a small number of patients and was ethnically homogeneous.

Conclusions

Our study demonstrated that when using Py as the neoadjuvant targeted therapy for HER2-positive BC, the tpCR and bpCR rates obtained from anthracycline-free chemotherapy regimens were not inferior to those obtained from anthracycline-containing chemotherapy regimens. In addition, the platinum-containing regimens and long-cycle taxane regimens were more likely to achieve pCR after omitting anthracyclines. Moreover, the ER status and HER2 status were independent influence factors of tpCR. Based on our study, HER2-positive BC patients with ER negative and HER2 IHC3+ are more recommended to adopt neoadjuvant chemotherapy regimens with Py. Nonetheless, further validation of above findings through large randomized controlled trials is warranted.

Availability of data and materials

The data generated in this study are not publicly available due to information that could compromise patient privacy but are available upon reasonable request from the corresponding author.

Declaration of competing interest

The authors declare that they have no competing interests.