Apatinib has demonstrated encouraging antitumor activity across a broad range of malignancies, including gastric, colorectal, and breast cancer, and good tolerability in phase I study conducted in our hospital [
18]. In a randomized, placebo-controlled phase II trial, which was also held in our hospital, the single-agent of apatinib showed potential efficacy in heavily pretreated metastatic gastric cancer [
19]. This phase II study was held to evaluate efficacy and safety of apatinib in patients with non-triple-negative MBC. Previous data suggested that the potential benefit from novel targeted agents was mediated through disease stabilization processes rather than tumor shrinkage. For instance, sorafenib has been reported to provide a statistically significant PFS and OS benefit in patients with hepatocellular and renal carcinomas without significant response rate improvement in two large randomized phase III trials [
21,
22]. Therefore, the primary end point of this study was designated as PFS. ORR and OS was the secondary endpoint. The median PFS of all 38 patients was 4.0 months (95% CI, 2.8 m – 5.2 m) and OS was 10.3 months (95% CI, 9.1 m – 11.6 m), which were similar to the results from another phase II study of apatinib in metastatic triple negative breast cancer (median PFS was 3.3 and OS was 10.6 months) [
20]. Of 36 patients eligible for efficacy analysis, ORR was 16.7% (6/36) and DCR was 66.7% (24/36). The surprise is 2 of the 6 CR/PR patients were HER2-positive and 1 of them was trastuzumab-pretreated. The results were encouraging for the efficacy seems superior to or at least comparable with that was reported in previous studies involved single-agent angiogenesis inhibitors. In a phase I/II trial of bevacizumab reported by Cobleigh et al., 75 heavily pretreated MBC patients were enrolled. The response rate was 9.3% and confirmed response rate was 7%. The median duration of confirmed response was 5.5 months (range, 2.3 to 13.7 months) [
23]. Two phase II studies investigating sorafenib in pretreated MBC got a TTP or PFS of 58 days (95% CI, 52 to 112 days) and 2.0 months (95% CI, 1.7 to 4.1 months), respectively. The response rate were 2% and 0, respectively [
15,
16]. Similarly, in a phase II study of sunitinib involving 64 pretreated MBC patients, the ORR was 11% and the median TTP was 10 weeks [
17]. The median overall survival in these trials were 43 weeks for bevacizumab [
23], 37 weeks for sorafenib [
15] and 38 weeks for sunitinib [
17]. It was therefore concluded that single-agent activity of angiogenesis inhibitors was limited and combination with standard chemotherapy was recommended. As a result, a series of four randomized, double-blind, placebo-controlled Phase IIb Trials were developed to Investigate the Efficacy of Sorafenib (TIES) when added to selected chemotherapies for HER2-negative MBC. In the SOLTI-0701 study, sorafenib plus capecitabine as first- or second-line significantly improved median PFS compared with placebo plus capecitabine (6.4 vs 4.1 months, HR = 0.58, 95% CI, 0.41–0.81, P = 0.001) [
24]. The AC01B07 study reported that the combination of sorafenib with gemcitabine or capecitabine in patients progressed during or after bevacizumab got a median PFS of 3.4 months and an ORR of 19.8% [
25]. Oppositely, sorafenib plus first-line paclitaxel did not significantly improve PFS (6.9 months for sorafenib vs 5.6 months for placebo, HR = 0.788, 95% CI, 0.558-1.112, P = 0.1715, 1-sided P = 0.0857) in the NU07B1 study [
26], nor did its combination with docetaxel and/or letrozole as first-line treatment in the FM-B07-01 study [
27]. Besides the older VEGF-TKIs of sorafenib and sunitinib, the more recently introduced VEGF-TKI of axitinib, which was able to inhibit VEGF receptors at subnanomolar concentrations, also didn’t improve TTP when combined with docetaxel in first-line MBC treatment compared with docetaxel plus placebo (8.1 v 7.1 months, HR = 1.24, 95% CI, 0.82 -1.87, 1-sided P = 0.156). However, in the subgroup analysis of patients who had received prior adjuvant chemotherapy, an improvement in TTP was observed (9.2 v 7.0 months, P = 0.043), suggesting the potential of axitinib to reverse chemotherapy resistance [
28]. Although the data from those studies above all indicated potential activity for VEGF-TKIs in combination with selected chemotherapies, phase III trials were necessary for confirmation. We hypothesized that VEGF-TKIs might be effective in breast cancer with high angiogenesis dependency and the molecular subtypes of breast cancer such as TNBC or non-TNBC was not a potential efficacy predictor. The most frequently observed AEs of apatinib of all grade in this study were hand-foot syndrome (52.6%), proteinuria (52.6%) and hypertension (42.1%), which were similar to those reported in the phase I study of apatinib in metastatic gastric cancer [
18]. Most AEs were mild to moderate (grades 1 to 2) in severity. 16.6% AEs were Grade 3 and no grade 4 toxicities were observed. Although one patient died within 28 days of last treatment and one died of intestinal obstruction after receiving 16 days of treatment of apatinib, the two deaths were both considered to be the result of disease progression. Hemotologic toxicities including neutropenia and thrombocytopenia were mild to moderate and no dose interruption or reduction was needed during the treatment. 73.7% patients experienced dose interruption and 47.4% received dose reduction during treatment because of non-hemotologic toxicities. 12.1% patients discontinued treatment due to an AE and the majorities due to disease progression. The mechanism of hypertension is thought to be the inhibition of VEGFR in arterial endothelial cells leading to decrease of the release of nitric oxide, which acts on arterial smooth muscle cells to cause vasodilation [
29]. Hypertension could be well controlled by using angiotensin receptor blocker (ARB, such as valsartan) with or without calcium antagonists (such as amlodipine) besides dose interruption or reduction. Hand-foot syndrome and proteinuria could also recover rapidly and be well tolerated after dose interruption or reduction. As a result, careful monitoring of toxicity and prompt dose interruption or reduction from 500 mg to 375 mg or 250 mg were essential during the treatment.