Trastuzumab, a humanized mAb, was the first drug approved for the treatment of HER2-positive metastatic BC (20). Trastuzumab mediates ADCC, inhibits the PI3K/AKT signaling pathway, blocks the G₁ phase of the cell cycle and inhibits DNA damage repair to serve an antitumor role (21,22). It is considered that the BBB allows substances of only small molecular weight (400-500 Da) (23) to pass through. Because of its large molecular mass (148,781 Da), trastuzumab is considered to be unable to cross the intact BBB. However, intracranial trastuzumab uptake and reduced BM have been observed in animal models of HER2-positive BCBM (24). The addition of trastuzumab to first-line treatment for HER2-positive metastatic disease has been documented to significantly increase the time to central nervous system (CNS) metastasis (25). The possible causes include disruption of the BBB during BM or after radiotherapy, in addition to increased vascular permeability due to neovascularization and the secretion of vascular endothelial growth factor (VEGF) (26). With tumor cell proliferation and new blood vessel formation, the blood-tumor barrier (BTB) forms between the tumor and blood vessels, where BTB permeability is significantly increased compared with the BBB. Therefore, some drugs that have difficulty passing through the BBB can enter brain metastases through intercellular bypass (27).

The registHER trial (28) previously demonstrated that patients with BM (n=258) who received trastuzumab after being diagnosed with CNS metastases had significantly longer OS compared with those who did not [n=119; 17.5 vs. 3.7 months, hazard ratio (HR)=0.25]. In addition, trastuzumab treatment (HR=0.33; 95% CI, 0.25-0.46; P=0.001) was found to be independently associated with a reduced risk of mortality after CNS metastasis. However, because of the low intracranial concentration of trastuzumab, these OS benefits were proposed to be due mainly to the long-term control of the extracranial disease (20).

At present, attempts have been made of using various biological techniques to increase drug permeability with the purpose of increasing the intracranial concentration of trastuzumab, such as combining MRI-guided focused ultrasound (MRgFUS) (29) with trastuzumab. MRgFUS was found to enhance drug uptake in 87±17% of sonicated voxels (>20% increase in standardized uptake value ratio), with ≤450% voxel-wise increase detected. In addition, nanoparticles can be used as carriers for macromolecular drugs to cross the BBB (30). However, the efficacy of nanoparticles remains in the experimental stage and there is no specific data.

Pertuzumab is a recombinant humanized mAb that can binds to subdomain II of the extracellular domain of HER2, blocking the dimerization of HER2 with HER3 and HER2 with HER1, which inhibits the downstream PI3K and MAPK pathways (31). It also exhibits a complementary mechanism with trastuzumab, resulting in dual obstruction when used in combination (32). According to the CLEOPATRA trial results (33), the time to CNS progression was found to be delayed with pertuzumab (15 vs. 11.9 months; P=0.0049) (34), which may be due to the superior control of extracranial disease with pertuzumab. In a retrospective study (35), the combination of pertuzumab and trastuzumab significantly improved the OS to 44 months, compared with that after other-HER2-targeted therapy (17 months) and no-HER2-targeted therapy (3 months) in patients with HER2-positive BCBM. In addition, the PATRICIA study (36) showed that pertuzumab combined with high-dose trastuzumab (6 mg/kg weekly) achieved a CNS objective response rate (ORR) of only 11% (95% CI, 3.0-25.4, median duration of response, 4.6 months). However, a large proportion of patients achieved clinical benefit at 4 (68%) and 6 months (51%). Notably, two patients had stable intracranial and extracranial disease for >2 years. In the final efficacy analysis based on the latest follow-up, the median CNS-progression-free survival (PFS) was found to be 4.6 months, whereas the systemic PFS was also 4.6 months and the median OS was 27.2 months (37). These results suggest that it is possible to further optimize the dose and regimen of mAbs to combat BCBM.

Margetuximab (MGAH22; margetuximab-cmkb) (38) is a human/mouse chimeric and Fc-optimized mAb against HER2. Margetuximab and trastuzumab share the same HER2 receptor-binding epitope. However, margetuximab exhibits increased binding capacity to CD16A (FcγRIIIA) and decreased binding capacity to CD32B (FcγRIIB), thereby enhancing the activation of innate and adaptive immune responses whilst maintaining the antiproliferative effect of trastuzumab. In the SOPHIA (NCT02492711) (39) clinical trial, a total of 71 patients (71/536) had BM at baseline. Margetuximab improved the primary PFS over trastuzumab [HR, 0.76; 95% CI, 0.59-0.98; P =0.03; median, 5.8 (95% CI, 5.5-7.0) months vs. 4.9 (95% CI, 4.2-5.6) months], whilst demonstrating an acceptable safety profile, making it an alternative option for late-line therapy. However, the overall OS analysis did not reveal the superiority of margetuximab over trastuzumab, where its intracranial response rate requires further exploration.

Lapatinib is a first-generation TKI for the treatment of HER2-positive BC and is a tyrosine kinase inhibitor of both HER1 and HER2. In 2013, a single-arm phase II multicenter study (LANDSCAPE) (44) evaluated the efficacy of lapatinib plus capecitabine in patients with HER2-positive BC who had not previously received WBRT. The CNS-ORR was calculated to be 57.1%, where the median PFS was 5.5 months. BM is reduced by >80% in 67% of patients. In addition, in the NRG Oncology-KROG/RTOG 1119(45) randomized trial, 143 patients were allocated to the following two groups: WBRT (37.5 Gy/3 weeks)/SRS (dose on the basis of lesion size) with or without lapatinib (1,000 mg per day for 6 weeks). The results reported that radiotherapy with lapatinib had a greater ORR (55 vs. 42%) at 4 weeks, suggesting that the combination may provide a short-term benefit. Therefore, the efficacy of lapatinib in the treatment of BM needs further clinical trial data for verification.

Pyrotinib is a small-molecule irreversible TKI that can target HER1, HER2 and HER4(46). It was approved in China in 2018 in combination with capecitabine for the treatment of patients with advanced or metastatic BC (46). The PERMEATE trial (47) was designed to examine the efficacy of pyrotinib plus capecitabine in two separate cohorts of patients with HER2-positive BCBM. In cohort A (patients who had not previously received local radiotherapy), the intracranial ORR was 74.6% (median follow-up duration was 15.7 months). In cohort B (patients with active BM who progressed after radiotherapy), the intracranial ORR was 42.1% (median follow-up duration was 15.7 months). The latest follow-up data (48) (median follow-up, 40.5 months) revealed that cohort A had a median PFS of 10.7 months (95% CI, 7.6-14.9) and a median OS of 35.9 months (95% CI, 25.1-not reached). This previous study confirms the efficacy of the pyrotinib-capecitabine regimen for the treatment of HER2-positive BM.

In addition, the combination of pyrotinib, capecitabine and radiotherapy have also been assessed in another study in patients with HER2-positive BC and BM. According to BROPTIMA (49), the 1-year CNS-PFS rate was 74.9% (95% CI, 61.9-90.7), the median CNS-PFS was 18.0 months (95% CI, 15.5-not reached) and the CNS-ORR was 85%, with 17 patients (42.5%) achieving complete response (CR) and partial response (PR). In addition to effectively controlling intracranial lesions, pyrotinib effectively reversed extracranial lesions, with an overall median PFS of 17.6 months (95% CI, 12.8-34.1). The safety of this combination was acceptable and it did not significantly impair central nervous system function.

Neratinib is an oral irreversible TKI of HER1, HER2 and HER4(50). The NEfERT-T study (51) compared neratinib with trastuzumab for the treatment of previously untreated HER2-positive metastatic BC (MBC), where the rates of symptomatic and progressive CNS recurrence were 8.3 and 17.3%, respectively (P=0.002). Although neratinib plus paclitaxel was not found to be superior to trastuzumab plus paclitaxel in terms of PFS for patients with HER2-positive MBC as a first-line treatment, it may delay the occurrence and decrease the incidence of CNS metastasis. In the randomized controlled phase III NALA trial (52), neratinib plus capecitabine was demonstrated to be superior to lapatinib plus capecitabine in terms of 1-year PFS (29 vs. 15%) for patients who received HER2-targeted therapies with two or more lines. In addition, neratinib provided superior control of CNS metastases (CNS-ORR, 26.3 vs. 15.4%) and can delay the occurrence of symptomatic BM requiring intervention (cumulative incidence, 22.8 vs. 29.2%). In the TBCRC022 Co3 phase II trial (53) investigating prior CNS-directed therapy and neratinib plus capecitabine in patients with BCBM, cohort A (who did not receive prior lapatinib) had a higher CNS-ORR rate compared with that in cohort B (who received prior lapatinib), with the CNS-ORR reported to be 49 vs. 33%. The median PFS was also different between the two cohorts (cohort A, 5.5 months; cohort B, 3.1 months). These findings suggest that neratinib exhibits a definite control effect on HER2-positive BM in patients treated with lapatinib.

Tucatinib is an oral and reversible TKI that is highly selective for the HER2 domain (54). According to the HER2CLIMB study (55), the risk of intracranial progression was reduced by 68% (HR=0.32; 95% CI, 0.22-0.48; P<0.0001) in patients with HER2-positive BCBM. The median OS of the tucatinib group was 6.1 months longer compared with that of the control group (18.1 vs. 12.0 months). The CNS PFS was 9.9 months (95% CI, 8.4-11.7) in the tucatinib group, which was also superior to that observed in the control group (4.2 months; 95% CI, 3.6-5.7). According to follow-up (56) data (median follow-up of 26.9 months), the tucatinib combination group had greater clinical benefits in terms of CNS PFS and intracranial ORR. This subgroup analysis further supported the importance of tucatinib combination therapy for patients with HER2-positive brain metastases.

This trial also included a specialized subgroup of patients with untreated BM who used tucatinib plus trastuzumab and capecitabine in place of radiation therapy (66 patients). The median CNS-PFS in the tucatinib group was 8.1 months. Although the sample size was small, this analysis revealed that the combination of tucatinib with trastuzumab and capecitabine may provide an OS benefit for patients with untreated BM, possibly even delaying the duration of radiation therapy, which may cause cognitive impairments (11).

T-DM1(58), which consists of trastuzumab and a cytotoxic component (the anti-microtubule drug DM1), exerts antitumor effects by inhibiting HER2 downstream signaling pathways, promoting ADCC, disrupting the microtubule network and causing apoptosis. Patients enrolled into the EMILIA study (59) were randomly allocated to the T-DM1 or lapatinib plus capecitabine groups. In addition, 45 and 50 patients had BM at baseline, respectively. In this subgroup, according to the RECIST 1.1 criteria (60), the median OS in the T-DM1 group was found to be 26.8 months, which was significantly superior compared with that in the lapatinib plus capecitabine group (12.9 months, HR=0.382; 95% CI, 0.18-0.80; P=0.0081).

The TH3RESA study (61) included patients with HER2-positive advanced BC who previously used trastuzumab and lapatinib. The subgroup analysis (RECIST V1.0) (62) revealed that T-DM1 treatment for patients with stable BM had a median OS of 17.3 months, which was superior compared with that of the investigator-selected treatment group (12.6 months, HR=0.62; 95% CI, 0.34-1.13), which was consistent with the beneficial trend noted in the overall population.

In another large-scale study, KAMILLA (63), among the 2,002 patients enrolled, 398 patients had BM at baseline. The median PFS was 5.5 months, whereas the median OS was 18.9 months (95% CI, 17.1-21.3). Among the 126 patients with measurable BM, the best overall response rate was 21.4%. Furthermore, 27 patients achieved stable disease (SD) for >6 months (assessed by RECIST V1.1). In total, 42.9% (54/126) patients had a >30% reduction in the sum of the maximum diameters of the target lesions. The clinical benefit rate was calculated to be as high as 42.9%. Therefore, notable antitumor activity was observed with T-DM1.

T-DXd (64) is comprised of a humanized mAb against HER2 linked to a DNA topoisomerase I inhibitor (Dxd) through a cleavable tetrapeptide linker. The amino acid sequence of the antibody partially overlaps with that of trastuzumab, but DXd is 10-fold more potent than SN-38, the active metabolite of irinotecan (65). Upon binding to HER2, T-DXd disrupts HER2 signaling and triggers ADCC (64). In addition, endocytosis-mediated cleavage of T-DXd inside cells releases DXd, which induces DNA damage and apoptosis (66).

DESTINY-Breast03(67) was the first head-to-head clinical study of ADC agents designed to compare the therapeutic effect of T-DXd with that of the second-line standard treatment T-DM1. A total of 43 and 39 patients with stable BM were enrolled into the T-DXd and T-DM1 groups, respectively. The median PFS for patients with stable BM treated with T-DXd was 15 months (95% CI, 12.5-22.2), which was significantly superior compared with that of patients with stable BM treated with T-DM1 (3 months, 95% CI=2.8-5.8). According to RECIST V1.1, the intracranial ORR of the T-DXd group was 63.9% (67), which was approximately twice as high as that of the T-DM1 group (33.4%). The intracranial CR rates of the T-DXd and T-DM groups were 27.8 and 2.8%, respectively, where that of the T-DXd group was ~10X higher compared with that of the T-DM1 group (68). One of the limitations was that the number of patients with stable BM was small, whereas patients with active BM were excluded from this trial.

The TUXEDO-1 trial (69) was initiated to fill the gap in data on the potential activity of T-Dxd in active BMs. The median PFS was 14 months, 13.3% (2/15) achieved CR, 60% (9/15) achieved PR and 20% (3/15) achieved SD. The best CNS-ORR was 73.3% (RANO-BM criteria) (70), which met the predefined primary study endpoint. The results indicated that T-Dxd not only exhibited clinically relevant activity in HER2-positive BC with active BM but also prolonged disease control despite the presence of BM. Therefore, these findings suggest that T-Dxd can be safely used to treat patients with HER2-positive BCBM to delay the initiation of local therapy. According to the latest final outcome analysis (71), the median follow-up was 26.5 months (95% CI, 23.5 months-not reached) and the median PFS was 21 months (95% CI, 13.3-not reached). The median OS was not reached (95% CI 22.2-not reached). Although TUXEDO-1 has several limitations, such as a single-center design and a small sample size, the long-term outcomes suggest that despite its large molecular size, T-DXd prolonged both intra- and extracranial disease control with acceptable tolerability whilst maintaining QoL function. The most frequent AEs were mild and moderate, such as fatigue (66.7%), nausea (46.7%) and anemia (46.6%).

Another multicenter retrospective real-world study, the ROSET-BM study (72), was performed to evaluate the efficacy of T-Dxd in patients with active BM and leptomeningeal carcinomatosis (LMC). The results from the total population revealed that the median PFS was 16.1 months, where the 1-year OS was 74.9%. Notably, among the 19 patients with LMC, the 1-year PFS and OS rates were 60.7 (95% CI, 34.5-79.1%) and 87.1% (95% CI, 57.3-96.6%), respectively. The results demonstrated sustained systemic and CNS disease control in patients with LMC.

Cohort 5 (patients with HER2-positive or HER2-low advanced or metastatic BC and untreated leptomeningeal carcinomatosis) of the DEBBRAH study (73) also showed positive activity in previously untreated HER2-positive patients with pathologically proven LMC, with no new safety concerns. Since the study remains at the initial stages, a longer follow-up will provide prospective data for this rare subgroup. In symptomatic patients, the sequence of systemic vs. local treatment may also have an impact on outcomes. The ongoing DESTINY-Breast12 study (74) (NCT04739761) will confirm the efficacy of T-DXd in patients with active BM in BCBM and will provide the greatest amount of information to date (n=250).

Novel HER2-targeted ADCs, such as SYD985, ARX788 and RC48, are being explored. ARX788 (Anviti) (75) is a locus-specific ADC drug that was independently developed in China and has been applied in phase III clinical research in the field of BC (76). It introduces para-acetyl-phenylalanine (pAF) into the HER2 mAb and binds to the cytotoxic tubulin inhibitor amberstatin 269 (AS269) to form a stable oxime bond. This structure minimizes untargeted toxicity caused by the shedding of cytotoxic drugs during circulation and effectively reduces the total amount of drug required for therapy (77). Upon endocytosis into cells, ARX788 then releases pAF-AS269 into lysosomes, which induces cell death by binding to microtubules (79). This agent is currently being explored in patients with HER2-positive advanced BM (NCT05018702).

RC48 (adastuximab) (79), which is comprised of the HER2-targeted antibody disitamab loaded with the toxin monomethyl auristatin E (MMAE) and the valine-citrulline linker, has more potent antineoplastic activity compared with T-DM1. RC48-ADC showed consistent efficacy in both the HER2-positive subgroup and the HER2-low expression subgroup (NCT02881138 and NCT03052634) (80). In addition, 2.0 mg/kg Q2W had a more favorable benefit-risk ratio compared with the other doses (81). In addition, in trastuzumab- and lapatinib-resistant xenograft tumor models in nude mice (79), this novel ADC drug has also shown superior antitumor activity compared with that by T-DM1, suggesting its potential as an improved therapy for HER2-positive BC.

Although the use of ADCs can significantly improve the survival outcome of patients, the duration of tumor remission or benefit of ADC monotherapy remains limited because of the emergence of drug resistance mechanisms (90). Therefore, combination strategies with other antitumor drugs have become an important direction for drug development. In another cohort from the TBCRC022 study (91), the efficacy of neratinib plus T-DM1 was evaluated in patients with previously untreated BM (cohort 4A) and patients without prior T-DM1 therapy who progressed after local therapy (cohort 4B). Cohort 4C included patients who had received T-DM1 after local therapy progression. The CNS-ORRs in cohorts 4A, B and C were 33.3, 29.4 and 28.6%, respectively (median follow-up time: Cohort 4A, 33 months; cohort 4B, 28 months; cohort 4C, 28 months). In another randomized, double-blind, phase III study, HER2CLIMB02(92), revealed the advantages of ADCs plus small-molecule tyrosine kinase inhibitors in HER2-positive advanced BC. The results indicated that in both patients previously treated with trastuzumab and taxane and patients with locally advanced/metastatic BC with a past history of BM, the median PFS of T-DM1 plus tucatinib was significantly longer compared with that of T-DM1 alone (9.5 vs. 7.4 months; HR=0.76). In addition, patients with BM (40% of all enrolled patients) also had a significant PFS benefit (median PFS, 7.8 vs. 5.7 months; HR=0.64), where no additional adverse events were observed in the combination group compared with T-DM1 alone. The combination of tucatinib plus T-DXd and tucatinib plus T-DM1 has been investigated in HER2CLIMB-04 (NCT04539938) (93) and CompassHER2 RD (NCT04457596) (94) patients, the data of which may provide novel opportunities for the treatment of patients with advanced BC and BM.

According to studies on the synergistic mechanism of HER2-targeted ADCs combined with immunotherapy, HER2-targeted therapy and immunotherapy involve complex crosstalk mechanisms. Anti-HER2 therapy promotes CD8+ T-cell infiltration into tumor tissue (95). The secretion of IFN-γ (96) by CD8+ T cells then enhances the anti-HER2-inhibiting effect on cell proliferation and upregulates programmed death-ligand 1 (PD-L1) expression. The HER2-targeted ADC SHR-A1811(97) for injection can bind to and endocytose HER2-expressing tumor cells, cleave toxins through proteases in tumor cell lysosomes and induce cell cycle arrest and apoptosis. Adebrelimab (98) is a humanized anti-PD-L1 mAb that can specifically bind to PD-L1 molecules to block the programmed cell death protein 1/PD-L1 pathway, which leads to tumor immune tolerance and reactivates the antitumor activity of the immune system to achieve tumor treatment. Currently, clinical trials of adebrelimab combined with SHR-A1811 (NCT05353361) in BC have been approved.

A novel targeted protein degradation system (99) has been recently designed based on phage-assisted continuous evolution technology. Degrons that can achieve targeted degradation, such as PROTAC, do not affect other non-target proteins and retain native genomic expression profiles (100), thereby minimizing interference with regulatory mechanisms that are critical to the natural biological functions of numerous proteins. In recent years, on the basis of the HER2-selective inhibitor tucatinib, a HER2 degrader was developed using PROTAC technology. The CH7C4(101) compound was obtained by optimizing the linker, E3 ligase ligand and binding site. CH7C4 effectively inhibits HER2-driven cancer cell proliferation through durable HER2 degradation and strong inhibition of downstream pathways, where its antitumor activity was stronger compared with that of tucatinib both in vitro and in vivo (101). This technology provides novel ideas for the development of new drugs for the treatment of HER2-positive BC.

CAR-T (102) cells are generated from T cells isolated from patients' peripheral blood and engineered in vitro to express synthetic receptors that recognize tumor antigens. CAR-T cells are subsequently cultured for expansion and infused back into patients (103). In the past, this type of immunotherapy was mainly used for treating hematologic tumors. Several clinical trials in different solid tumors are underway around the world (104), including for BC (NCT04650451, NCT03740256 and NCT03696030). Experimental data (105) have demonstrated that trastuzumab-resistant tumors can be effectively eliminated by HER2-CAR-T cells, indicating that the clinical use of trastuzumab-derived HER2-specific CAR-T cells represent an option for the treatment of trastuzumab-resistant tumors. In addition, other experimental data (106) have demonstrated that HER2-CAR-T cells containing the 4-1BB costimulatory domain exhibit superior tumor targeting ability, reducing the T-cell exhaustion phenotype and enhancing the proliferative capacity compared with those containing the CD28 costimulatory domain. Local intracranial delivery of HER2-CARs has shown potent antitumor activity in orthotopic xenograft models and in regional intraventricular delivery (106,107), which is undoubtedly worthy of further research.