In 2020, the incidence of breast cancer (BC) surpassed lung cancer and was ranked first, accounting for 11.7% of all cancer, with about 2.3 million new cases, worldwide (1). With the BC treatment development, the frequency of central nervous system (CNS) metastases has steadily increased because BC patients survive long enough to be at risk of developing CNS metastasis (2). About 7% of metastatic BC will develop brain metastases (BMs) (3), including parenchymal and leptomeningeal disease, accounting for roughly 17% of all BMs, and the second major cause of BMs after lung cancer (4, 5). Risk of BM is variable across BC subtypes. BM is majorly observed in triple-negative BC (TNBC) and human epidermal growth factor receptor 2 (HER2)-positive BC. In comparison with the hormone receptors (HR) positive subtype, the risk of BM development in HER2 positive and TNBC is 2-5 times higher (6, 7). Although the advanced BC treatment has remarkably prolonged the patient’s survival, the prognosis of BC brain metastasis (BCBM) is still substandard. The median patients’ survival in parenchymal and leptomeningeal disease was 3 to 23 and 3 to 4 months, respectively (8, 9). The survival time for HER2+ patients is reported to be the longest, while that of TNBC patients is considered the least (10). In a large multi-center study, the median overall survival (OS) after the diagnosis of BCBM with HR+/HER2+ was 18.9 months, with HR–/HER2+ was 13.1 months, with HR+/HER2– was 7.1 months and for triple-negative was 4.4 months (11). Due to the substandard prognosis and lack of efficient treatment strategies, BCBM represents a unique and challenging clinical problem.

The brain is the most complex and unique organ in the body, and because of this, there are differences in treatment. The brain microenvironment is composed of two different components, parenchyma and leptomeninges (12). The brain parenchyma comprises cells that are not present anywhere else in the body, which include astrocytes, oligodendrocytes, microglia, and neuronal cells. The leptomeninges is mainly filled with the circulating cerebrospinal fluid (CSF) produced by the choroid plexus (13). The blood-brain barrier (BBB) is the main gatekeeper of the CNS (13). BBB is a unique neurovascular unit consisting of a continuous segment of non-porous blood vessels and interacts with parietal, immune, glial, and nerve cells to carefully modulate the movement of ions, molecules, and cells between the brain and the blood (14, 15), thereby protecting the CNS from pathogens, toxins, injury, inflammation, and diseases, while also providing a barrier for drugs delivery to the brain (16). Another important factor that produces drug resistance during BM treatment is the BBB efflux transporter systems, which reversely transport and prevent drug penetration into CNS. These include P-glycoprotein, BC-resistance protein, multidrug resistance-associated proteins, etc. (17). The formation of metastatic tumors may partially disrupt the BBB, allowing it to become more permeable. However, it is still not sufficiently or homogeneously permeable for effective drug therapy (18).

Currently, the main treatment strategies for BCBM include local (surgical resection and radiotherapy) and systemic treatments (19). In the study by Minniti et al, it has been demonstrated that BM patients who received multifraction (3× 9 Gy) stereotactic radiosurgery after surgery had an improved local control rate of 91% at 12 months (20). Systemic treatments are important for controlling CNS diseases and improving patients’ survival (19). Although heavy literatures on BBB’s effect and anti-cancer drugs’ efficacy on BCBM is deficient, our knowledge about the effect of systemic treatments on BCBM is rapidly increasing. This review focuses on the currently available system treatment drugs for BCBM, such as chemotherapy, target therapy, endocrine therapy, immunotherapy, and novel therapies that may be available in the future ( Figure 1 ).

HER2+ BC accounts for approximately 20% of all BC cases (21). HER is a group of reversible tyrosine kinase receptors, comprising four members: epidermal growth factor receptor (EGFR)/HER1, HER2, HER3, and HER4. These have crucial functions in tumorigenesis and help tumor cells escape anti-tumor immunity. Anti-HER2-targeting drugs can inhibit the kinase activity of these HER receptors and prevent further cancer cell survival and proliferation (22). HER2-targeting drugs involve tyrosine kinase inhibitors (TKIs), monoclonal antibodies, and antibody-drug conjugates (ADCs). TKI is a small molecule and comprises lapatinib, neratinib, tukatinib, etc. have a considerably lower molecular weight, allowing them a more efficacious penetration through the BBB (21). Table 1 lists the characteristics of studies reporting on outcomes related to BMs in patients with HER2-positive BC.

Monoclonal antibody drug, such as trastuzumab, was the first authorized targeted therapy against HER2+ BC. In the registHER study, trastuzumab treatment after CNS metastatic diagnosis significantly improved OS statistically (treatment vs. no trastuzumab treatment, 17.5 vs. 3.8 months) (23). Trastuzumab also has inconsistent responses in intracranial and extra-cranial tumors, as it cannot entirely cross the BBB to play a preventive role (39). A meta-analysis of HER2+ BC patients (n=9020) showed that CNS metastasis accounted for a greater proportion of the first relapse site in patients receiving trastuzumab treatment (24). Among HER2+ patients receiving trastuzumab adjuvant therapy, the CNS metastasis incidence as the first site of disease recurrence was 2.56% (95% CI 2.07%-3.01%) than in those who did not receive adjuvant trastuzumab 1.94% (95% CI 1.54%-2.38%) (24). Pertuzumab acts on a different HER2 site, and it prevents HER2/HER3 dimerization (40) which can lead to significant PI3K-Akt activation (main signaling pathway in BC cells) (41). When combined with trastuzumab, it can play a complementary role and provides efficient treatment for HER2+ BC patients. However, this dual-target therapy fails to show an advantage in BMs (42). In the phase II PATRICIA investigation, the CNS ORR for pertuzumab plus high-dose trastuzumab was 11% (95% CI 3-25) (25). In early studies, the trastuzumab–pertuzumab dual-target therapy combined with other agents indicated some activity in BCBM (26, 27, 43). A retrospective observational investigation which included 264 HER2+ BCBM patients revealed that dual HER2-blockade and taxanes treatment produced an ORR of 52.4% in baseline BM patients (26). In the phase III PHEREXA trial, after pertuzumab was added to capecitabine + trastuzumab, the PFS markedly elevated in the subgroup of HER2+ baseline BM patients (n = 53; HR 0.29, 95% CI 0.15–0.60) (27). All in all, trastuzumab and partuzumab, as macromolecular monoclonal antibodies, have limited their ability to pass the blood-brain barrier, but their combination with other chemotherapy drugs has increased their intracranial activity to some extent.

Lapatinib inhibits EGFR and HER2 and is authorized to be given with capecitabine as a combined therapy for metastatic BC (44). The CNS ORR of mono lapatinib therapy is approximately only 6% (28), while the CNS ORR of lapatinib + capecitabine in 13 BCBM patients who prior received trastuzumab and radiation therapy is 38% (95% CI, 13.9-68.4) (45). At the same time, a single-arm phase II clinical trial showed that in untreated BCBM patients, the CNS ORR of lapatinib + capecitabine could reach 65.9% (29). Lapatinib + capecitabine is also associated with less incidence of CNS at the first progression. A phase III randomized trial indicated that in comparison with capecitabine monotherapy, the CNS involvement cases at first progression were fewer (2% vs 6%, P = 0.045) in combined therapy (46). Lapatinib combined with fractionated radiotherapy may be useful against HER2+ BCBM in the tumor xenograft model (47). In a Phase I investigation, lapatinib with WBRT had a higher CNS ORR (79%) than that of WBRT alone (36%) in 28 BCBM patients (30).

Neratinib irreversibly inhibits HER2, HER1, and HER4 (48), and may pass through the intact BBB by inhibiting ATP-binding cassette B1 transport function which is one of the dominant efflux transporters in the BBB (49, 50). In a phase II trial of neratinib combined with capecitabine treatment in refractory HER2+ BCBM patients, 33% and 49% of the patients with and without previous lapatinib treatment achieved CNS ORR, respectively (31). In a randomized clinical trial of previously untreated metastatic HER2+ BC, 8.3% and 17% of patients in the neratinib + paclitaxel and trastuzumab + paclitaxel groups experienced symptomatic or progressive CNS recurrence, respectively (32), demonstrating its preventive effect on the CNS metastasis of BC.

Pyrotinib is an irreversible TKI that targets HER1, HER2, and HER4 (51). A phase II clinical trial (PERMEATE trial) showed the intracranial ORR was 74.6% in a radiation-naive population with pirotinib + capecitabine (cohort A) (n=59) and 42.1% in cohort B (n=19) included patients who had progressive disease after radiotherapy (33). The median PFS in cohort A was 11.3 months and 5.6 months in cohort B (33). PERMEATE research adds strong medical evidence for drug treatment of BM, especially for patients with new BM. However, the efficacy of this regimen still needs more randomized controlled trials for further verification.

Tucatinib is a selective HER2-targeting TKI, and has fewer side effects than other TKIs due to its lack of strong EGFR inhibition (10). In 2020, Tucatinib is authorized by the Food and Drug Administration (FDA) for HER2+ BCBM treatment. A randomized clinical trial (HER2CLIMB trail) was conducted where tucatinib was given with trastuzumab+ capecitabine as a combination therapy to patients who were initially treated for HER2+ metastasis BC. In BM patients, the estimated 1-year PFS in the tucatinib and placebo groups was 24.9% (95% CI, 16.5 to 34.3) and 0%, respectively. The median PFS in the tucatinib and placebo groups was 7.6 months (95% CI 6.2-9.5) and 5.4 months (95% CI 4.1-5.7), respectively. The CNS ORR was 40.6% (95% CI 35.3-46.0) in tucatinib group compared with 22.8% (95% CI 16.7-29.8) in placebo group (P<0.001). However, the side effects such as the incidences of diarrhea and hepatic injury (grade 3 or higher) were more frequent than those in the control group (34).

Trastuzumab emtansine (T-DM1), an ADC of trastuzumab and the cytotoxic drug emtansine, a maytansine derivative and microtubule inhibitor, has been aproved for the treatment of HER2+ metastatic BC patients after taxane and trastuzumab therapy, and for the adjuvant therapy of early BC HER2+ patients with the residual invasive disease after neo-adjuvant taxane and trastuzumab therapy (43, 52). In the EMILIA study, the patients with baseline CNS metastases had OS of 26.8 months in the T-DM1 arm, compared with 12.9 months in the capecitabine + lapatinib arm (HR=0.38, P=0.008). PFS in the two treatment arms was similar (5.9 vs. 5.7 months; P=1.0) (35). In a KAMILLA single-arm research, in the 126 measurable BCBM patients, CNS ORR was 21.4% (95% CI 14.6–29.6), whereas, in the 398 baselines BCBM patients, the median PFS and OS were 5.5 months (95% CI 5.3–5.6) and 18.9 (95% CI 17.1–21.3), respectively (36). In short, T-DM1 may be effective in treating HER2+ intracranial lesions.

Another ADC is trastuzumab deruxtecan (DS-8201), comprising a human HER2 antibody, a new enzyme-cleavable linker, and a topoisomerase I inhibitor payload. Compared with T-DM1, the antibody-drug ratio of DS-8201 is higher (8 vs 3–4) (37). The DESTINY-Breast01 trial, evaluating DS-8201 in HER2+ metastatic BC patients with previously treated with T-DM1, revealed a median PFS of 18.1 months (95% CI 6.7–18.1) in the BM subgroup (37). In DESTINY-Breast03, in the 82 BCBM patients, DS-8201 had better efficiency than T-DM1. The median PFS was 15.0 months (95%CI 12.5-22.2) in the DS-8201 group and 3.0 months (95%CI: 2.8-5.8) in the T-DM1 group (HR=0.25; 95%CI 0.31-0.45), and the ORR of DS-8201 group and T-DM1 group were 67.4% and 20.5%, respectively (38, 53).

To summarize, the available data indicate remarkable intracranial efficacy from treatments for HER2-positive BCBM, such as TKIs, monoclonal antibodies, and ADCs. However, some problems need to be solved, for example, treatment-induced diarrhea and hepatic injury may be limiting factors and require appropriate surveillance, and the optimal administration time and sequence of each agent remain uncertain.

The patients suffering from HR+ BC are less likely to develop BMs than patients with other subtypes of BC (6). The effects of hormone therapy specifically on BMs are still unclear (54). Concentrations of tamoxifen and its metabolites in the BM tumor and brain tissue were 46 times higher than those in serum, suggesting that tamoxifen might be clinically beneficial (55). Furthermore, in preclinical models, estrogen promoted BMs by altering polarity and suppressing the phagocytic activity of M2 microglia, whereas tamoxifen blocked its polarization and enhanced its phagocytic ability, thereby inhibiting BMs (56). Other hormone therapies, such as megestrol acetate and letrozole, only a small number of case reports have documented the responses of hormonal therapy in BCBM patients, and the efficacy has not been confirmed in large-sample clinical trials (54, 57–60).

The growth of HR+ BC cells depends on cyclin D1, which activates cyclin-dependent kinases 4 and 6 (CDK4 and CDK6), thereby inducing the G1-S phase transition and entering the cell cycle. CDK4/6 inhibitors are reliable treatment options for HR+ BC patients with extra-cranial diseases (10). In preclinical models, CDK4/6 inhibitor abemaciclib has better central permeability than other existing CDK4/6 inhibitors (61, 62), which can reach the therapeutic level in human BM (62). In the JPBO study, abemaciclib was given to HR+/HER2- BCBM patients, and the results showed that 25% of patients did achieve clinical benefit (CR, PR, or SD>6 months), but definite intracranial ORR was only 5.6% (63). The real effectiveness of CDK4/6 inhibitors in treating BMs is not verified yet.

Everolimus is a kinase that acts selectively to inhibit the mammalian target of rapamycin (mTOR), and can easily penetrate the CNS of a mouse model (64). However, the intracranial response in a phase II study of everolimus, trastuzumab, and vinorelbine for the treatment of progressive HER2+ BCBM was only 4% (65). But in the phase Ib/II trial assessing the effect of everolimus, lapatinib, and capecitabine combined therapy for HER2+ BCBM, the CNS ORR was 28% (66), highlighting the need for exploring better treatment modalities for everolimus.

In the future, compared with the systemic treatment of patients with BM, prevention of BM from the primary tumors seems to be a more important clinical goal. How to accurately screen the high-risk population with BM from BC patients requires more clinical research to explore a reliable prediction model. With the rise of liquid biopsy technology represented by circulating tumor cells (CTCs), it provides strong support for the detection of circulating brain-tropic cancer cells before their extravasation (114). The gene signature of CTCs associated with BCBM has revealed the up-regulation of Notch signaling (124). Notch targeted therapy may specifically reduce the incidence of BM in BC.

BM is an important clinical source of morbidity and mortality in patients with metastatic BC (43). Local interventions are the pillar of BM management (19). Systemic treatments are often used to complement local strategies to achieve optimal control of CNS diseases (19). At present, with the emergence of various new drugs, some systematic therapies have shown promising clinical results. The best results published to date have been obtained in patients with HER2+ BCBM, especially in new BM, for whom currently used combinations of chemotherapy and anti-HER2 therapy have shown certain efficacy, with particularly impressive results obtained with pirotinib + capecitabine and trastuzumab deruxtecan. However, patients with BM from HR+ BC or TNBC lack effective medical options currently, PI3K inhibitors, FASN inhibitors and immunotherapy are promising therapeutic candidates. In the coming years, the results of ongoing clinical trials with combinations of immunotherapy, radiotherapy and targeted therapies may provide better treatment options for BCBM patients.

Conceptualization, YZ. Writing – original draft, QC and JX. Writing – review and editing. QC, JX, YM and JW. Supervision, CL and YZ. All authors contributed to the article and approved the submitted version.