The role of capecitabine-based neoadjuvant and adjuvant chemotherapy in early-stage triple-negative breast cancer: a systematic review and meta-analysis

Background The role of capecitabine in neoadjuvant and adjuvant chemotherapy for early-stage triple-negative breast cancer (TNBC) is highly controversial. Our meta-analysis was designed to further elucidate the effects of capecitabine on survival in early-stage TNBC patients and its safety. Methods PubMed, Embase, and papers presented at several main conferences were searched up to December 19, 2019, to investigate capecitabine-based versus capecitabine-free neoadjuvant and adjuvant chemotherapy in TNBC patients. Heterogeneity was assessed using I2 test, combined with hazard ratios (HRs) and odds ratios (ORs) with 95% confidence intervals (CI) computed for disease-free survival (DFS), overall survival (OS), and over grade 3 adverse events (AEs). Results A total of 9 randomized clinical trials and 3842 TNBC patients were included. Overall, the combined capecitabine regimens in neoadjuvant and adjuvant chemotherapy showed significantly improved DFS (HR = 0.75; 95% CI, 0.65–0.86; P < 0.001) and OS (HR = 0.63; 95% CI, 0.53–0.77; P < 0.001). In subgroup analysis, there were improvements in DFS in the groups with addition of capecitabine (HR = 0.64; 95% CI, 0.53–0.78; P < 0.001), adjuvant chemotherapy (HR = 0.73; 95% CI, 0.63–0.85; P < 0.001), and lymph node positivity (HR = 0.62; 95% CI, 0.44–0.86; P = 0.005). Capecitabine regimens were related to higher risks of diarrhea (OR = 2.88, 95% CI 2.23–3.74, P < 0.001), stomatitis (OR = 2.01, 95% CI 1.53–2.64, P < 0.001) and hand–foot syndrome (OR = 8.67, 95% CI 6.70–11.22, P < 0.001). Conclusion This meta-analysis showed that neoadjuvant and adjuvant chemotherapy combined with capecitabine significantly improved both DFS and OS in early-stage TNBC patients with tolerable AEs. There were benefits to DFS in the groups with the addition of capecitabine, adjuvant chemotherapy, and lymph node positivity. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-021-07791-y.

Background "Triple-negative" [negative for estrogen receptor (ER−), progesterone receptor (PR−), and human epidermal growth factor receptor 2 (HER-2)] breast cancers (TNBCs) account for approximately 15 to 20% of all breast cancers and have the characteristics of heterogeneity, aggressiveness, and poor prognosis [1][2][3]. Although programmed cell death 1 (PD1), programmed cell death ligand 1 (PD-L1) inhibitor, and poly (ADP-ribose) polymerase (PARP) inhibitor have been shown to be effective in the neoadjuvant phase [4][5][6][7], chemotherapy is still the backbone of treatment by neoadjuvant and adjuvant therapy [8]. National Comprehensive Cancer Network Guidelines and the St. Gallen International Expert Consensus recommended standard regimens containing anthracycline and paclitaxel for early-stage breast cancer patients [9,10]. However, even if the chemotherapy regimen is effective, the 10-year risk of relapse of early TNBC is still up to 20-40% [11][12][13]. Therefore, it is crucial to explore new adjuvant and neoadjuvant chemotherapy drugs and regimens, improve the treatment efficacy, and translate any progress into survival benefits.
Capecitabine is one of the most widely studied drugs in the neoadjuvant and postoperative adjuvant setting of TNBC. Recently, several randomized clinical trials (RCTs) have evaluated the clinical value of combined treatment with capecitabine for TNBC, but the outcomes of these studies were still controversial [14][15][16][17][18][19][20][21][22]. Seven meta-analyses summarizing the role of capecitabine in breast cancer adjuvant and neoadjuvant chemotherapy have been performed, six of which included all subtypes of patients, not just TNBC patients [13,[23][24][25][26][27][28]. Some meta-analyses estimated disease-free survival (DFS) of TNBC in a subgroup, and only three reported overall survival (OS) [13,26,28], but the DFS and OS of TNBC results were controversial. Although a meta-analysis was recently performed to investigate the role of capecitabine in early-stage TNBC [13], it included only seven studies; it appeared that most of the participants in the research by Zhang et al. were hormone receptor-and/ or HER-2-positive, while the meta-analysis dealt with all patients having TNBC. In addition, none of the meta-analyses mentioned above included the latest data of two studies (CBCSG-010 and CIBOMA) that particularly focused only on TNBC, and also did not distinguish neoadjuvant chemotherapy from adjuvant chemotherapy as it appeared that patients could not benefit from the addition of capecitabine to neoadjuvant chemotherapy [29]. Our meta-analysis enlarges the sample size and refines the subgroup analysis possible to make the conclusion more robust.
To further clarify the role of capecitabine in TNBC, in the present study, we conducted a meta-analysis to investigate the effect of capecitabine as neoadjuvant and adjuvant chemotherapy on survival in TNBC patients.

Methods
This analysis was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRIS MA) guidelines [30].

Search criteria
Randomized controlled trials (RCTs) were identified using a computerized search of the databases PubMed and Embase (up to December 2019), using the MeSH terms "breast cancer," "breast neoplasms," "triple-negative breast cancer," "triple-negative breast neoplasms," "capecitabine," "Xeloda," "neoadjuvant chemotherapy," and "adjuvant chemotherapy." The search was limited to articles written in English. We also searched the proceedings of the last 10 years of annual meetings at the American Society of Clinical Oncology, European Society for Medical Oncology, and San Antonio Breast Cancer Symposium, including relevant unpublished studies. We used the following criteria: (i) phase II or III RCTs; (ii) TNBC patients who received neoadjuvant or adjuvant chemotherapy including capecitabine in the experimental arm and chemotherapy without capecitabine in the control arm; and (iii) studies reporting DFS and/or OS risk ratios. We excluded those with no reported hazard ratios (HRs) of 95% confidence intervals (CI) for DFS or recurrence-free survival (RFS) and OS in TNBC, as well as non-RCT, meta-analysis, and retrospective studies.

Data extraction
Two investigators (XH and FZ) independently extracted data of the RCTs by a search strategy, and then proceeded to aggregate the results; divergence was resolved by consensus. The following information was extracted: trial name, authors' names, update year, trial design, trial phase, number of TNBC patients, median follow-up, hormone receptor and human epidermal growth factor receptor 2 (HER-2) statuses, and chemotherapy regimens. The HRs and 95% CI of the efficacy measures for DFS and OS were extracted where available, and grade 3-5 drug-related adverse events (AEs) were also extracted. For studies with multiple reported outcomes, the latest reported data were used. The Cochrane Collaboration's tool for assessing risk of bias was used to assess the quality and potential bias of studies (Table S1).

Statistical analysis
The combined HRs of each of DFS and OS were weighted and combined by the generic inverse variance method [31]. Pooled estimates of odds ratios (ORs) were computed for grade 3-5 drug-related AEs. Subgroup analysis was conducted according to (i) adding and replacing capecitabine in chemotherapy, (ii) neoadjuvant and adjuvant chemotherapy, or (iii) lymph node positivity and negativity. Heterogeneity was assessed using I 2 test [32]. When P < 0.10 or I 2 > 50%, the random effects model was used, and when P > 0.10 or I 2 < 50%, the fixed effects model was used [33]. The possibility of publication bias was estimated by observing a funnel plot (Figure S1). P < 0.05 was considered significant. All statistical analyses were performed using Stata 12.0 software.

Characteristics of the studies
A total of 508 records were retrieved based on the search terms; after excluding 499 irrelevant records, nine potentially eligible RCTs were considered [14][15][16][17][18][19][20][21][22], one of which was a conference abstract [21] ( Figure S2). However, the data for the FinXX trial are based on RFS, not DFS. Because the definition of DFS in the other trials resembles that of RFS in the FinXX trial, we combined the RFS of FinXX with the DFS of the other trials to perform a joint analysis. A total of 3842 TNBC patients were included in this meta-analysis. In two of the trials, the entire cohort of patients had TNBC [19,21], while in the other seven trials TNBC patients were one of the subgroups. The main research features are shown in Table 1.

DFS and OS
The HRs of nine RCTs for both DFS and OS were pooled. From the available data, there were 1757 cases of involving the addition or replacement of capecitabine in the capecitabine group and 1425 cases of chemotherapy without capecitabine. The combined capecitabine regimens in neoadjuvant and adjuvant chemotherapy were associated with significantly improved DFS (HR = 0.75; 95% CI, 0.65-0.86; P < 0.001) with low heterogeneity (P = 0.192, I 2 = 28.4%) (Fig. 1). They also significantly improved OS (HR = 0.63; 95% CI, 0.53-0.77; P < 0.001) with low heterogeneity (P = 0.702, I 2 = 0) (Fig. 2).

DFS subgroup analysis Addition and replacement of capecitabine
There were five [14,[17][18][19]21] and four [15,16,20,22] studies involving the addition or replacement of capecitabine to neoadjuvant or adjuvant chemotherapy, respectively. From the available data of TNBC, there were 1373 capecitabine combined and 1356 capecitabine-free cases in the group with the addition of capecitabine, and 384 capecitabine combined and 357 capecitabine-free cases in the group with the replacement of capecitabine. DFS was significantly improved in the group with the addition of capecitabine (HR = 0.64; 95% CI, 0.53-0.78; P < 0.001) with high heterogeneity (P = 0.573, I 2 = 0), but not in the group with the replacement of capecitabine (HR = 0.88; 95% CI, 0.73-1.08; P = 0.225) with low heterogeneity (P = 0.359, I 2 = 6.8%) (Fig. 3).

Grade 3-5 adverse event analysis
Considering that the influences of adverse events related to hormone receptor and HER-2 statuses were small, we used the overall population's AE data to perform this meta-analysis. In total, 4682 of 37,675 (12.43%) patients were in the capecitabine group and 5029 of 37,276 (13.49%) were in the capecitabine-free group. There were four grade 3-5 hematological AEs, three grade 3-5 gastrointestinal AEs, and two grade 3-5 other AEs. Figure 6 shows an overview of the safety profile for grade 3-5 AEs in the capecitabine and capecitabine-free chemotherapy groups.

Discussion
To the best of our knowledge, this is the most comprehensive meta-analysis based on the most recent studies which evaluates the efficacy and safety of neoadjuvant and adjuvant chemotherapy regimens combined with capecitabine for TNBC. With nine studies and a total of 3842 TNBC patients in this meta-analysis, the neoadjuvant and adjuvant chemotherapy combined with capecitabine for TNBC patients not only improved DFS, but also improved OS. In subgroup analysis, benefits to DFS were identified in the groups with the addition of capecitabine, adjuvant chemotherapy, and lymph node positivity, but not in those with the replacement of capecitabine, neoadjuvant chemotherapy, and lymph node negativity. Capecitabine regimens were related to higher risks of grade 3-5 AEs for diarrhea, stomatitis, and HFS, but lower risks of leucopenia, neutropenia, and thrombocytopenia, while there were no significant differences in the AEs neutropenic fever, fatigue, and vomiting compared with the case for capecitabine-free regimens.
Capecitabine has been used for treating advanced breast cancer and is efficient after advancement in an anthracycline-or taxane-based setting [34][35][36]. Previous studies also showed that capecitabine may have specific benefits for TNBC patients [37,38]. Considering the efficacy of capecitabine for treatment in a metastatic setting and TNBC, speculation has been reported about its potential role in early TNBC and several large clinical trials on this issue have been carried out. However, these studies reported conflicting results and few studies focused on TNBC. The overview of seven meta-analyses can be found in Table S2. A meta-analysis included all subtypes of breast cancer patients, and TNBC patients were analyzed in subgroup analysis only [23]. But in our metaanalysis, we included only patients with TNBC and subgroup analysis included the addition/replacement of capecitabine, adjuvant/neoadjuvant chemotherapy, and lymph node status, we also increased the analysis of AEs of stomatitis. Recently, a meta-analysis from Li et al. [13] reported that TNBC patients might benefit from the addition of capecitabine in the adjuvant settings, however, this meta-analysis not only included TNBC patients, but included hormone receptor-and/or HER-2positive patients. In our article, we analyzed TNBC patients, and two latest study results (GEICAM and CBCSG-010). Although several meta-analyses based on limited studies and patients summarized the role of capecitabine in neoadjuvant and adjuvant chemotherapy for breast cancer [13,26,28], the results regarding DFS and OS in TNBC were still controversial. More recently, three studies (CALGB 49907, GEIC AM/2003-11_CIBOMA/2004-01, and CBCSG-010) provided updated results, two of which only included TNBC patients treated with capecitabine adjuvant or neoadjuvant chemotherapy. Thus, it is necessary to reevaluate the efficacy of capecitabine in neoadjuvant and adjuvant chemotherapy. Our results revealed the significant extension of DFS and OS by capecitabine neoadjuvant and adjuvant chemotherapy in early-stage TNBC patients. For patients with positivity for hormone receptor or HER-2, the effects of adjuvant treatment can be intensified by inhibiting ovarian function and prolonging endocrine therapy [39][40][41][42][43] or by other anti-HER-2 drugs including pertuzumab, neratinib, and TDM1 [44][45][46]. Based on our results, capecitabine might become an intensive treatment protocol for patients with TNBC.
Overall, five studies (FinXX Trial, US Oncology 01062, CREATE-X, CIBOMA 2004/01, CBCSG-010) and four studies (GEICAM/2003-10, GAIN, Gepar TRIO, CALGB49907) have been performed on the addition or replacement of capecitabine in neoadjuvant or adjuvant chemotherapy, respectively. The E1199, ABC, and S0221 trials suggested that the inclusion of anthracyclines in addition to taxanes and alkylator chemotherapy is effective for early-stage TNBC patients [47][48][49]. The chemotherapy regimens were all based on anthracyclines and/ or taxanes in the group with the addition of capecitabine in our meta-analysis, so DFS can be further significantly improved. However, the replacement of capecitabine with anthracycline and/or taxane will compromise the survival of patients.
This meta-analysis also revealed that there was a survival benefit of using capecitabine only in adjuvant chemotherapy, but not in a neoadjuvant setting. Two meta-analyses [29,50] showed no significant improvement in the rate of pathologic complete response (pCR) in neoadjuvant treatment combined with capecitabine. Achieving pCR after neoadjuvant treatment can significantly improve DFS in TNBC patients [51][52][53]. The negative relationship between capecitabine and pCR in a neoadjuvant setting might account for capecitabine not improving the DFS of TNBC patients. Nevertheless, there was a significant improvement in the rate of DFS in the group with adjuvant chemotherapy combined with capecitabine, which may be related to the long-term effects of capecitabine on dormant tumor cells, activating anti-cancer immunity or antiangiogenic activity [54][55][56].
Only two studies (CIBOMA 2004/01, CBCSG-010) reported the relationship between lymph node status of TNBC patients and neoadjuvant or adjuvant chemotherapy including capecitabine. Our meta-analysis showed that survival benefits were only conferred in lymph node-positive patients. Lymph node metastasis is one of the risk factors for TNBC patients [19]. Patients who did not reach pCR in the CREATE-X trial [18] and received capecitabine could benefit from capecitabine as intensive treatment. Thus, capecitabine might be more efficacious for high-risk TNBC patients. Similarly, capecitabine only improved the survival of lymph node-positive patients, but not lymph node-negative ones.
We also evaluated the AEs associated with capecitabine as neoadjuvant and adjuvant chemotherapy. Our meta-analysis showed that capecitabine-based neoadjuvant and adjuvant chemotherapy increased the risks of diarrhea, stomatitis, and HFS, but reduced the risks of grade 3-5 hematological AEs such as leucopenia, neutropenia, and thrombocytopenia. Physicians should be aware of the possibility of several capecitabine-related AEs in clinical practice, such as HFS, diarrhea, and stomatitis, which might lead to the discontinuation of treatment. The neoadjuvant and adjuvant capecitabine combined chemotherapy was negatively related to the risk of grade 3-5 hematological toxicities, which may be associated with the replacement of capecitabine monotherapy in some capecitabine groups. Nevertheless, the toxicity of capecitabine was acceptable and easily manageable in the patients overall.
Our meta-analysis has some limitations. First, this is a meta-analysis based on the literature but not on individual patient data. Thus, biased findings might have been obtained regarding the treatment efficacy and toxicity. Second, the studies included in this meta-analysis have several potential differences, such as the definition of TNBC and the chemotherapy regimens. In some studies, the rate of cases of TNBC as defined by ER/PR expression was less than 10% [14], while in other studies it was less than 1% [19]. The most used chemotherapy regimens were based on anthracyclines, taxanes, and cyclophosphamide, but there were still some differences among the studies, such as the dose of chemotherapeutics, and one study also used antiangiogenic drugs. This heterogeneity should be taken into account when interpreting our findings. Third, the results of one study were only provided in conference abstract form, but not as a full report. Some information was incomplete, such as patient age, and dose and duration of chemotherapy. Fourth, the types of AEs reported in the included studies differed, and some AEs were not reported in several studies. Considering that the influences of AEs related to hormone receptor and HER-2 status were small, we used the overall population's AE data to perform this metaanalysis, but the comprehensive effect of AEs may be biased compared with the actual results.

Conclusion
In conclusion, our meta-analysis showed that neoadjuvant and adjuvant chemotherapy combined with capecitabine significantly improved both DFS and OS in earlystage TNBC patients. DFS was significantly improved in the groups with the addition of capecitabine, adjuvant chemotherapy, and lymph node positivity, but not in those with the replacement of capecitabine, neoadjuvant chemotherapy, or lymph node negativity. Capecitabine regimens were related to higher risks of grade 3-5 AEs, namely, diarrhea, stomatitis, and HFS, but negatively correlated with the risks of leucopenia, neutropenia, and thrombocytopenia.