Serum miR-222-3p as a Double-Edged Sword in Predicting Efficacy and Trastuzumab-Induced Cardiotoxicity for HER2-Positive Breast Cancer Patients Receiving Neoadjuvant Target Therapy

Background: We aimed to explore whether the expression of serum miR-222-3p might contribute to early prediction of therapeutic response, clinical outcomes, and adverse events for HER2-positive breast cancer patients receiving neoadjuvant therapy (NAT). Methods: A total of 65 HER2-positive breast cancer patients receiving NAT were analyzed. The concentration of serum miR-222-3p was detected by quantitative real-time PCR. Logistic regression analysis was used to identify the association of serum miR-222-3p with pathological complete response (pCR). The relationship of serum miR-222-3p with disease-free survival (DFS) and overall survival (OS) was examined via log-rank test and Cox proportional hazards analysis. The ordered logistic regression was applied to evaluate the association between serum miR-222-3p and adverse events. Results: The miR-222-3p low group was more likely to achieve pCR [odds ratio (OR) = 0.258, P = 0.043]. The interaction between miR-222-3p and presenting Ki67 level was also detected for pCR (OR = 49.230, Pinteraction = 0.025). The miR-222-3p low group was correlated with superior DFS (P = 0.029) and OS (P = 0.0037). The expression of serum miR-222-3p was the independent protective factor for trastuzumab-induced cardiotoxicity (P < 0.05) and anemia (P = 0.013). Conclusions: Serum miR-222-3p is the potential factor to predict pCR, survival benefit and trastuzumab-induced cardiotoxicity for HER2-positive breast cancer patients receiving NAT.


INTRODUCTION
Neoadjuvant chemotherapy (NAC) is increasingly used in primary breast cancer patients, which not only aims to improve the operability or breast-conservability, but also serves as a good platform for in-vivo tests of various drugs (1)(2)(3). Trastuzumab, an anti-HER2 monoclonal antibody, is well exemplified in the treatment of HER2-positive breast cancer (4). Manifold data have demonstrated that the addition of trastuzumab to NAC significantly improves the pathological complete response (pCR) rates and thereby results in survival benefit (5)(6)(7)(8). However, a majority of HER2positive breast cancer patients still failed to achieve pCR or even progressed despite trastuzumab-based neoadjuvant therapy (NAT) (9)(10)(11)(12). Given the aggressive biological behavior of HER2positive breast cancer, it hints at a demand to identify potential biomarkers to predict its response to NAT.
On the other hand, adverse events, especially trastuzumabinduced cardiotoxicity, also accompany during or after NAT. The overall incidence of cardiotoxicity was reportedly 3-7% for trastuzumab monotherapy, 13% for trastuzumab with paclitaxel, and as high as 27% for trastuzumab with anthracycline (13,14). Unfortunately, few reliable biomarkers could help to predict the trastuzumab-induced cardiotoxicity.
Liquid biopsy, as a minimally invasive test, has developed dramatically in recent years. MicroRNAs (miRNAs) belong to a class of noncoding, regulatory, single-stranded RNAs, which have been reported to contribute in early detection of treatment efficacy and adverse reaction (15)(16)(17)(18)(19)(20)(21)(22)(23). Our previous research showed that the expression level of miR-222-3p in serum declined after surgery and was an independent prognostic factor for disease-free survival (DFS) in breast cancer (24). Basic studies revealed that miR-222-3p could upregulate HER2 signaling pathway in fulvestrant-resistant breast cancer cells and inhibit the autophagy of cardiac myocytes in mice (25)(26)(27). Previous studies have revealed that miR-222-3p was associated with immune invasion and immune resistance in a variety of tumors. Overexpression of miR-222-3p was found to enhance the resistance of tumor cells to tumor infiltrating lymphocytes (TILs) by down-regulating the expression of intercellular cell adhesion molecule-1 (ICAM1) in melanoma, which resulted in ipilimumab (anti-cytotoxic T lymphocyte-associated antigen-4 antibody) resistance in patients with melanoma (28). On the other hand, Ying et al. found that in epithelial ovarian cancer, cancer cell-derived exosomes with high contents of miR-222-3p transferred to the tumor-associated macrophages (TAM) and then induced their polarization to the M2 phenotype via SCOX3/STAT3 pathway (29). The transformation of TAM from M1 to M2 phenotype predicted poor prognosis (30) and contributed to the resistance to anti-HER2/Neu treatment in breast cancer (31). However, it still remains illdefined whether serum miR-222-3p can serve as a potential biomarker for predicting the response to NAT in HER2positive breast cancer patients as well as their trastuzumabinduced cardiotoxicity.
On these premises, we hypothesized that the expression of serum miR-222-3p might contribute to early prediction of therapeutic response, clinical outcomes and adverse events for HER2-positive breast cancer patients receiving NAT.

Study Procedure
All the enrolled HER2-positive breast cancer patients came from two neoadjuvant clinical trials registered as SHPD001 (NCT02199418) and SHPD002 (NCT02221999) in ClinicalTrials.gov. The SHPD001 and SHPD002 trials were verified and authorized by the Independent Ethical Committee of Renji Hospital, Shanghai Jiaotong University. Each patient signed written informed consent.
The eligibility criteria for these two neoadjuvant trials included women aged ≥18 and ≤70 years old with locally advanced invasive breast cancer (T2-4 or N1-3) confirmed independently by two pathologists based on World Health Organization (WHO) classification. All the patients received paclitaxel 80 mg/m 2 on day 1, 8, 15 and 22 and cisplatin 25 mg/m 2 on day 1, 8 and 15 every 4 weeks for 4 cycles. For HER2-positive patients, concurrent weekly trastuzumab was given at a loading-dose of 4 mg/kg, followed by maintenance dose of 2 mg/kg, on day 1 for 16 weeks. However, 6 HER2-positive breast cancer patients couldn't afford to receive trastuzumab. In the SHPD002 trial, hormone receptor (HR)-positive patients were randomly assigned to receive preoperative endocrine therapy or not. Endocrine therapy hereinto referred to gonadotropin releasing hormone agonist for premenopausal women and letrozole for postmenopausal counterparts, concurrently with NAC. Tumor assessment was performed every 2 months by physical examination, mammary magnetic resonance imaging (MRI) and ultrasonography. All the patients were required to measure the left ventricular ejection fraction (LVEF) at baseline and every 3 months thereafter. Adverse events were graded according to Common Terminology Criteria for Adverse Events (CTCAE) 4.0. After completion of NAT, the patients underwent surgery. Up until April 2018, 65 HER2-positive breast cancer patients were available for this analysis from these two trials. The results were reported and the analysis was devised according to the Tumor Marker Prognostic Report Recommendation (REMARK) guidelines (32,33).

Study Outcome
The primary outcome of SHPD001 and SHPD002 was pCR, which was defined as the absence of tumor in the breast tissues and auxiliary lymph nodes removed at the time of surgery (ypT0 ypN0). The secondary outcomes of SHPD001 and SHPD002 included DFS, overall survival (OS) and adverse events. DFS was defined as the time from surgery until first occurrence of locoregional relapse, contralateral breast cancer, distant metastasis, second primaries and death from any cause. OS was defined as the time from surgery until death from any cause.

Pathological Examination
Estrogen receptor (ER), progesterone receptor (PR), HER2 and ki67 were evaluated by immunohistochemistry (IHC) in paraffin-embedded tumor samples from core needle biopsy. ER or PR positive was defined as ≥10% of stained cells.

Sample Collection
Before NAT, 5 mL peripheral blood was collected in the coagulation-promoting tubes and let stand at 4 • C for at least 1 h. Then it was centrifuged at 1,000 × g for 10 min at 4 • C to spin down the blood cells. The supernatants were centrifuged at 12,000 × g for 10 min at 4 • C to completely remove cellular components. Subsequently, the serum samples were divided into 300 ul/tube in RNA-free EP tubes and stored at −80 • C until use (24).  (0.006 < 2 − Ct ≤ 0.053) or high [2 − Ct > 0.053 (at the 75th percentile)]. The ordered logistic regression was applied to evaluate the association between serum miR-222-3p and adverse events. The relative and absolute drop of LVEF from baseline were respectively calculated when the LVEF measured at two time points or more were available. The relative drop of LVEF from baseline (rLVEF) was defined as (LVEF min -LVEF baseline )/LVEF baseline , and the absolute drop of LVEF from baseline (aLVEF) as LVEF min -LVEF baseline . Since grade 3/4 LVEF decrease occurred in only a minority of patients, either rLVEF or aLVEF was divided by its quartile. The statistical analysis was performed by STATA Statistics SE 14 (Stata Corp LP, College Station, TX, USA). All tests were two-tailed and P < 0.05 was considered statistically significant.

The Association Between the Expression of Serum miR-222-3p and the Clinicopathological Characteristics
The baseline characteristics of all the patients are listed in Table 1. The expression of serum miR-222-3p was associated with presenting clinical N stage (P = 0.016) and death (P = 0.046). The miR-222-3p low group tended to experience less DFS events (P = 0.066). Significant correlation failed to be discovered between serum miR-222-3p and other clinicopathological characteristics.

DISCUSSION
This study, to the best of our knowledge, is the first to report the association between the expression of serum miR-222-3p and the response to NAT in the HER2-positive breast cancer patients. This is also the first time to report the relationship between the serum miR-222-3p expression level and adverse events.  Serum miRNAs were identified as biomarkers to diagnose or predict prognosis in various carcinomas (34,35). Some basic studies demonstrated that tumor can release miRNAs into circulation and serum miRNAs exist in a remarkably stable form protected from endogenous RNase activity (36,37), while others indicated the presence of miRNA-enriched exosomes secreted by non-cancer cells such as adipose tissue macrophages and mesenchymal stromal cells (38)(39)(40). Therefore, certain miRNAs, with cancer-, tissue-or organ-specific functions, might produce endocrine, paracrine or autocrine effect (41-44) on breast cancer patients, directly or indirectly regulating their response to NAT. Our data reveal that higher level of serum miR-222-3p was associated with inferior pCR rate, which might be attributed to trastuzumab-resistance. So far, two important pathways might contribute to trastuzumab resistance. Firstly, the activation of phosphatase and tension homologs (PTEN) predicts inhibitory effect of trastuzumab, whereas PI3K pathway activation through PTEN loss and PIK3CA mutation confers trastuzumab resistance in breast cancer (45)(46)(47)(48)(49). Secondly, PTEN loss leads to SRC hyperactivation, and SRC inhibition appears to overcome trastuzumab resistance (50)(51)(52)(53). On the other hand, miR-222 was reported to promote adriamycin or tamoxifen resistance through PTEN/Akt pathway in breast cancer (54,55). Consequently, miR-222-3p may also induce trastuzumab resistance by modulating the PTEN/PI3K/Akt pathway and the expression of SRC, which prompts further investigation. We observed the poor prognosis in the patients with higher level of serum miR-222-3p. An increasing number of researches have demonstrated that miR-222 promotes epithelialto-mesenchymal transition (EMT) (26,56,57), G1/S transition of cell cycle (58) and cell proliferation (59,60) in breast cancer. Therefore, the expression level of miR-222-3p is responsible for the invasion and metastasis of breast cancer. The prognostic value of miR-222 has also been clarified not only in breast cancer (24,61) but also in many other malignancies including pancreatic cancer, bladder cancer and glioblastoma (62)(63)(64)(65)(66). Furthermore, the anti-miR-222 might improve the chemosensitivity and survival, which provides a novel strategy for breast cancer management (67).
Trastuzumab administration is associated with an increased risk of cardiovascular adverse events (68)(69)(70). However, the underlying mechanism remains suspended. Trastuzumabinduced cardiotoxicity, regarded as a type II adverse drug reaction, commonly cause reversible dysfunction without cell loss (71). It generally presents as asymptomatic LVEF decline (72)(73)(74). Moreover, LVEF decline frequently improves with drug interruption and resuming trastuzumab after recovery is often feasible (71,74). This study showed that the impairment of LVEF was associated with the expression level of serum miR-222-3p in this study. MiR-222 has been found to participate in many physiological and pathological processes in the cardiovascular system (75). Additionally, miR-222 is necessary for cardiomyocyte growth induced by exercise and is sufficient to protect against adverse cardiac remodeling after ischemic injury (76,77). Serum miR-222-3p may play an endocrine role similar to that of hormones, reaching every organ, tissue or cell in the body with blood circulation. When miR-222-3p acts on cancer cells, it plays a role in carcinogenesis and drug resistance, and when it acts on cardiac myocytes, it plays a role in protecting the heart. Therefore, the overexpression of miR-222-3p might prevent the heart from trastuzumab-induced injury.
As far as we know, this is the first time that patients with lower expression level of serum miR-222-3p have been reported prone to anemia after receiving NAT. However, the molecular mechanism has not been clarified. The progressively down-regulating of miR-222-3p was observed during the normal erythropoiesis (78). In addition, miR-222-3p inhibited the expansion of erythroblasts and hematopoietic differentiation via down-regulated c-kit expression and miR-222-3p decreased the expression of BLVRA and CRKL to suppress erythroid differentiation (79)(80)(81). Consistent with previous basic studies, our data also showed that the patients with high miR-222-3p level have lower baseline hemoglobin level (OR = 0.07, 95% CI 0.005-0.991, P = 0.049). These results suggested that the erythroblasts may be more resistant to chemotherapy in the high miR-222-3p group due to the lower proliferative activity and conversely patients with low level of miR-222-3p might be prone to anemia.
The limitations of this study were as follows. Firstly, the sample size was relatively small. Large-scale studies are needed to confirm the predictive and prognostic value of miR-222-3p. Secondly, some follow-up data are missing on adverse events. Thirdly, as the serum samples were collected before NAT in this retrospective study of prospective trials, it was not possible for us to evaluate the correlation of miR-222-3p changes at different time points during or after NAT with various outcomes including response, survival and safety, which warrants further research.
In conclusion, our study revealed that serum miR-222-3p is the potential factor to predict pCR, survival benefit and trastuzumab-induced cardiotoxicity for HER2-positive breast cancer patients receiving NAT. However, the logic behind it is still open to investigation.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Independent Ethical Committee of Renji Hospital, Shanghai Jiaotong University. The patients/participants provided their written informed consent to participate in this study.