Neoadjuvant Docetaxel plus Carboplatin Versus Epirubicin plus Cyclophosphamide Followed by Docetaxel in Triple-negative, Early-stage Breast Cancer (NeoCART): Results from a Multicenter, Randomized Controlled, Open-label Phase II Trial

Background: Previous studies have shown that the addition of carboplatin to neoadjuvant chemotherapy regimens improved the pathologic complete response (pCR) rate in patients suffering from triple-negative breast cancer (TNBC). However, no studies have assessed the effects of the combination of docetaxel and carboplatin without anthracycline with taxane- and anthracycline-based regimens. Methods: The NeoCART study was designed as a multicenter, randomized controlled, open-label, phase 2 trial to assess the ecacy and safety of docetaxel combined with carboplatin with taxane- and anthracycline-based neoadjuvant chemotherapy in untreated stage II-III TNBC. All eligible patients were randomly assigned, at a 1:1 ratio, to the experimental DCb group (docetaxel plus carboplatin for six cycles) or the EC-D group (epirubicin plus cyclophosphamide followed by docetaxel). In both groups, pCR (ypT0/is ypN0) was evaluated as the primary outcome. Results: Between September 1, 2016, and December 31, 2019, 93 patients from 6 participating centers were randomly assigned, and 88 patients were evaluated for the primary end-point (44 patients in the DCb group and 44 patients in the EC-D group). In the primary end point analysis, 27 patients in the DCb group achieved a pCR (61.4%, 95% CI 47.0-75.8), and 17 patients in the EC-D group achieved a pCR (38·6%, 95% CI 24.3-53.0); with a difference of 22.8% (odds ratio 2.52, 95% CI 2.4-43.1; p non-inferiority=0.004), non-inferiority was met, and the DCb regimen was conrmed as superior to the EC-D regimen (p=0.044, superiority margin of 5%). At the end of the 37-month median follow-up period, overall survival and event-free rates were equivalent in both groups. The grade 3/4 adverse events in the DCb group included


Background
Studies show that patients with TNBC have poor prognosis due to the lack of targeted therapy. Despite advances in the treatment for primary breast cancer, cytotoxic chemotherapy is still the therapeutic base for TNBC.
Neoadjuvant therapy is usually recommended for TNBC. The bene ts of neoadjuvant therapy include reducing the tumor size for breast-conserving surgery, avoiding axillary lymph node dissection, making inoperable tumors operable, and obtaining an in vivo evaluation of the tumor's chemosensitivity. Taxaneand anthracycline-based neoadjuvant regimens have become a standard treatment for TNBC, leading to longer overall survival (OS) and event-free survival (EFS) in patients who achieve a pathologic complete response (pCR) 1 . The main focus of researchers has consistently been to improve the pCR rate of TNBC.
Platinum attacks cancer cells by inducing double-stranded DNA breaks, and TNBC may be sensitive to platinum 2 . Previous studies have shown a signi cantly improved pCR rate in patients with TNBC after adding carboplatin to neoadjuvant chemotherapy regimens 3,4,5 . Due to the long-term cardiotoxicity caused by anthracycline, several studies have explored the e cacy of neoadjuvant taxanes plus carboplatin regimens in TNBC and achieved satisfactory pCR rates 6, 7 . However, no studies have assessed the effects of the combination of docetaxel and carboplatin without anthracycline versus taxane-and anthracycline-based neoadjuvant regimens on the pCR rate.
The NeoCART study was designed to assess the e cacy and safety of docetaxel plus carboplatin versus epirubicin plus cyclophosphamide followed by docetaxel neoadjuvant chemotherapy in TNBC.

Patient Population
The following criteria were used to enroll patients: Eastern Cooperative Oncology Group (ECOG) performance status of 0-1; age ≥ 18 years, cytologically or histologically con rmed nonin ammatory invasive TNBC; a clinical stage of II-III (T1cN1-2 or T2-4N0-2) and previously untreated. The diagnosis of TNBC was performed in line with the College of American Pathologists guidelines and American Society of Clinical Oncology (ASCO). According to immunohistochemistry results, progesterone receptor (PR) and estrogen receptor (ER) positivity rates were less than 1%, the immunohistochemistry score for human epidermal growth factor receptor type 2 (Her2) staining was 0 or 1+, and uorescence in situ hybridization was utilized that no Her2 gene ampli cation. Adequate hematology, renal and liver function, normal heart function and negative pregnancy tests were analyzed for females with reproductive potential. Imaging tools (MRI, mammography or ultrasound) were used to measure the largest tumor diameter upon which T staging was determined. Staging evaluation comprised CT scanning for chest and abdomen and/or abdominal sonography and radionuclide bone imaging. Patient with metastatic disease or who had received anticancer treatment were not quali ed. Patient with a history of malignancy at another site (except for basal cell or squamous cell carcinoma of the skin and cervical carcinoma in situ) that had been fully treated were quali ed if no disease was present for more than ve years.

Study Procedures
The NeoCART study was designed as a multicenter, parallel group randomized controlled, open-label, phase 2 trial. The treatment allocation list was created, and randomization was performed centrally at the leading research center of Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences. All eligible patients were randomly assigned at a 1:1 ratio to the experimental DCb group (docetaxel (75 mg/m 2 administered intravenously every 3 weeks) plus carboplatin (AUC 6 mg/mL per min, intravenously every 3 weeks) for six cycles) or the EC-D group (epirubicin (90 mg/m 2 ) plus cyclophosphamide (600 mg/m 2 ), both administered intravenously every 3 weeks for four cycles, followed by docetaxel (100 mg/m 2 ) administered intravenously every 3 weeks for four cycles). The treatment regimen is shown in Fig. 1. Granulocyte colony-stimulating factors (G-CSF) use for prophylaxis of febrile neutropenia was permitted according to the ASCO guideline 8 . The hematological and biochemical indexes were evaluated every cycle. The dose of chemotherapy agents were reduced due to certain adverse events.
After surgery, those who exhibited residual disease were treated with capecitabine (1000-1250 mg/m 2 ) given two-times a day for 1-14 days and cycled every 21 days for a total of 6-8 cycles 9 .
At 3 and 8 weeks after the last neoadjuvant chemotherapy treatment, de nitive surgery was performed.
The type of breast surgery (mastectomy or breast-conserving surgery) and axillary treatment (sentinel lymph node biopsy or axillary lymph node dissection) were determined by the treating surgeon.
The pCR rate, which was de ned as the absence of invasive tumor cells in the breast and axilla (ypT0/is ypN0), was measured as the primary end-point. The pCR rate was determined by a local pathologist. Secondary endpoints included EFS and OS assessed by an investigator, breast-conserving surgery rates and treatment-related toxicities were also examined. The National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.0) was employed to assess the level of toxicity.

Data Collection and Analysis
Each number of subjects was randomly assigned to each group. Patients were randomized by means of a permuted block randomization scheme using an interactive response system (IxRS). Intention-to-treat methods were used and included in the analysis of all patients who started treatment.
The primary e cacy analysis was a non-inferiority comparison of the pCR rate between the DCb group and the EC-D group, using of a prespeci ed non-inferiority margin of -5% as the absolute difference between the two groups. If non-inferiority was met, the superiority of the DCb regimen was to be tested 10 .
Based on the ndings of previous studies 6 , we assumed that 30% of the patients in the EC-D group and 55% of the patients in DCb group would achieve a pCR. With these assumptions, we estimated that we would need to enroll 82 patients (randomized 1:1) considering a one-sided α value of 0.025 for the hypotheses with 80% power.
The nal date of data acquisition for this study was December 31, 2019. Data were analyzed with SPSS statistics version 22 (IBM Corporation). This study was registered at ClinicalTrials.gov (number NCT03154749).

Study Approval
The study was approved by the Research Ethics Committee of Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, and the protocol was reviewed by the responsible ethics committee is each center. This study conformed to the Declaration of Helsinki. All participants provided written informed consent.

Patient Characteristics
Between September 1, 2016, and December 31, 2019, 93 patients from 6 participating centers in China were randomly assigned (47 patients to the DCb group and 46 patients to the EC-D group). The intentionto-treat analysis was performed for all participants except ve who did not agree to receive treatment. Thus, 88 patients were evaluated for the primary end-point (44 in the DCb group and 44 in the EC-D group). Figure 2 shows the consort study owchart.
The demographic information of 88 subjects who received treatment was presented in Table 1. Patient and treatment features were balanced in both groups. The median age of the 88 patients was 50 years (range 19-69 years). In total, 1.1%, 35.3%, and 63.6% of patients had histological grades of I, II, and III, respectively. A total of 76.1% of patients had T1 or T2 tumors, and 56.8% of the axillary lymph nodes were positive. Among the 88 patients, 41 received a germline BRCA status test, and 5 (12.2%) had deleterious mutations. In the clinical evaluation before operation, 1 patient (2.3%) in the DCb group had stable disease (SD) and 1 patient (2.3%) had progressive disease; 4 patients (9.0%) in the EC-D group had SD. The proportions of breast-conserving surgery in the DCb and EC-D groups were 36.4% and 38.6%, respectively (p = 0.826).  (Fig. 5B).
Patients achieving pCR showed markedly higher EFS and OS as relative to subjects with residual disease.

Treatment Delivery and Toxicity
In total, 88.6% (39 of 44) and 97.7% (43 of 44) of patients in the DCb and EC-D groups, respectively, completed the scheduled treatment. Seven (15.9%) of 44 patients needed a decrease in carboplatin dosing from the initial dose of AUC 6 to AUC 5, and 1 patient (2.3%) needed a second reduction to AUC 4.

Discussion
Typically, TNBC is characterized by lack of Her2, PR, and ER. Due to the lack of other effective treatment, cytotoxic chemotherapy is the main therapy for TNBC. TNBC responds well to cytotoxic chemotherapy and has a higher pCR rate after neoadjuvant chemotherapy. Patients who obtained a pCR could achieve prolonged EFS and OS 1 . However, the clinical progression of TNBC is dismal relative to non-TNBC [11][12][13][14][15] . Therefore, the direction of current research is aimed at improving the pCR rate of neoadjuvant chemotherapy to improve the prognosis.
Over time, taxane-and anthracycline-based neoadjuvant regimens have become a standard treatment for TNBC. Some phase 2 clinical trials have shown that adding platinum to taxane-and anthracycline-based neoadjuvant regimens can signi cantly improve the pCR rate. In a GeparSixto trial, more patients achieved a pCR rate of 53% in the group combining carboplatin with paclitaxel plus non-pegylated liposomal doxorubicin and bevacizumab relative to those treated without carboplatin 4  analyzed the e cacy of docetaxel plus carboplatin (DCb). Through combined analysis of two cohorts, they found that the total pCR rate and residual cancer burden (RCB) 0 / 1 rate of the DCb regimen were 55% and 68%, respectively 7 . The pCR rate of this DCb regimen seems to be numerically higher than that of the taxane-and anthracycline-based neoadjuvant regimens reported in several trials 5,16,17 . As a result, National Comprehensive Cancer Network (NCCN) guidelines have added weekly paclitaxel plus carboplatin or docetaxel plus carboplatin regimens for select patients with TNBC in the preoperative setting 18 . Recently published NeoSTOP study was con rmed that the pCR and RCB 0 + 1 rates of carboplatin plus docetaxel (CbD) were similar to carboplatin plus paclitaxel followed by doxorubicin plus However, there is no study comparing the DCb regimen with the taxane-and anthracycline-based neoadjuvant regimens.
The NeoCART study is a multicenter prospective randomized controlled study to assess 6 cycles of DCb with epirubicin plus cyclophosphamide followed by docetaxel. Docetaxel combined with carboplatin showed a higher pCR rate (61.4%) with an absolute difference in the pCR rate of 22.8%. Our study supports the results of previous neoadjuvant chemotherapy studies on TNBC. In our study, the pCR rate was 38.6% in the control group, which is similar to the retes of 28%-46% reported in several trials of taxane-and anthracycline-based neoadjuvant regimens 3,5,16,17 .
This multicenter, prospective, randomized controlled study showed that 6 cycles of DCb chemotherapy is tolerable and 88.6% of patients completed all 6 cycles of treatment. In the analysis of adverse reactions, the use of carboplatin resulted in increased thrombocytopenia, but most patients experienced grade 1/2 adverse events; patients in the EC-D group may need a longer treatment time and are more likely to experience neutropenia.
A previous meta-analysis con rmed that patients with TNBC who achieved a pCR after neoadjuvant treatment had prolonged EFS and OS compared those who did not achieve a pCR 1 . According to the results of previous studies, the survival bene t of carboplatin is not clear. Our study showed that 6 cycles of DCb can signi cantly improve the possibility of achieving a pCR in patients with early TNBC and that the toxicity is controllable, while the rate of EFS and OS were similar to the taxane-and anthracyclinebased group.
In the BRCA status detection performed in our study, 5 patients (12.2%) had deleterious mutations, which was consistent with the results of previous studies 3,7,20 25 . Additionally, the preferred recommended regimens in the NCCN guidelines were dose-dense doxorubicin/cyclophosphamide followed by weekly paclitaxel or every 2 weeks (ddAC-T). However, a head-to-head comparison is need between the EC-D and ddAC-T regimens for their effectiveness as neoadjuvant chemotherapy. The evidence for ddAC-T in the NCCN guidelines originates from the CALGB 9741 trial, which focuses on adjuvant chemotherapy 26 . In the EBCTCG meta-analysis, only 2583 (6.9%) cases were included from neoadjuvant chemotherapy studies 25 . In addition, according to the results of the ECOG 1199 study, the e cacy of docetaxel administration every 3 weeks was the same as that of weekly paclitaxel 27 .
In summary, as assessed with the regimen of taxane-and anthracycline-based neoadjuvant therapy, docetaxel combined with carboplatin showed a higher pCR rate, a low incidence of controllable treatmentrelated adverse events, and the similar survival rates. Docetaxel combined with carboplatin is an effective and promising neoadjuvant chemotherapy regimen for TNBC. The results need to be veri ed in a phase III study, and longer follow-up data are needed for the survival results.

Conclusions
Docetaxel plus carboplatin showed a higher pCR rate than that of the taxane-and anthracycline-based regimens. There was no signi cant difference in the breast-conserving rate, EFS and OS between the two treatment arms. In addition, docetaxel plus carboplatin had a low incidence of controllable treatmentrelated adverse events. This study provides an effective regimen option for neoadjuvant therapy of triplenegative breast cancer.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests Figure 1 Schema of randomized phase II NeoCART trial.

Figure 2
The consort study owchart. DCb, docetaxel plus carboplatin; EC-D, epirubicin plus cyclophosphamide followed by docetaxel.  Effect of the two treatment regimens overall and in subgroups Event-free survival (A and C) and overall survival (B and D) according to treatment groups and pCR status. *pCR, pathological complete response