The human epidermal growth factor receptor 2 (HER2) is a tyrosine kinase receptor critically involved in the carcinogenesis of the mammary gland [1]. It is a member of the HER family of tyrosine kinase receptors, including three other members [epidermal growth factor receptor (EGFR), HER3, HER4], widely interconnected into a signaling network with important implications in breast oncogenesis and clinical behavior of breast tumors [2–6]. The study of HER2 oncogenic role and the development of drugs targeting the HER2 receptor have revolutionized breast oncology in the last decades. Indeed, about 15–20% of breast cancers (BCs) harbor amplifications of the ERBB2 oncogene, which encodes for HER2, and this alteration leads to a more aggressive behavior of disease compared to HER2-negative (HER2^–) cases [7, 8]. However, HER2 blockade with trastuzumab and several other compounds has significantly improved the prognosis of these patients. Indeed, in the early setting, a substantial percentage of patients are cured by the addition of (neo)adjuvant HER2-blockade to standard treatment [9–11], whereas in the advanced setting the overall survival (OS) of patients has been significantly extended with the introduction of anti-HER2 agents, possibly leading to a cure for a minority of them [12, 13]. In an attempt to further improve outcomes in this subgroup of patients, multiple pharmacological strategies are being investigated, with some recent successes and with a multiplicity of promising compounds that may redesign treatment algorithms in the years to come.

The purpose of this review article is to recapitulate the landscape of novel anti-HER2 compounds under investigation, as well as explore the evolution in treatment paradigms determined by the emergence of such agents.

Trastuzumab is the cornerstone of treatment for patients with HER2⁺ early BC (eBC). According to seminal studies about adjuvant treatment of HER2⁺ eBC, the addition of trastuzumab to standard adjuvant chemotherapy halves the risk of recurrence, with a 10% absolute improvement in long-term disease-free survival (DFS) and a 9% increase in 10-year OS [9, 14, 15]. In these trials, trastuzumab was administered concurrently with a taxane-based chemotherapy and continued thereafter to complete 1 year of treatment. Several trials tested longer (i.e. HERA trial [16]) or shorter durations (i.e. SOLD [17], Short-HER [18], E2198 [19], FinHER [20], PHARE [21], Hellenic trial [22], Persephone [23]). Among them, only the Persephone trial was able to show the non-inferiority of the 6-month regimen. However, a recent meta-analysis including 6 large trials testing trastuzumab de-escalation, a shorter duration of adjuvant trastuzumab was statistically noninferior to its 1-year administration and resulted in lower rates of cardiotoxicity [24]. Overall, a shorter duration of trastuzumab could be discussed for patients with low-risk HER2⁺ eBC and/or for patients at high risk of toxicity, while all the others should receive the standard 12-month regimen.

As for other BC subtypes, a neoadjuvant strategy is usually preferred to the adjuvant one [25], except for very small tumors (T < 2 cm, clinically node-negative). The phase III NOAH trial first showed a strong association between pathologic complete response (pCR) in the breast and axilla and long-term outcomes [26], subsequently confirmed by further neoadjuvant trials [27]. Dual HER2-targeting with pertuzumab added to chemotherapy plus trastuzumab further increased the pCR rate [11, 28], and led to pertuzumab approval by both EMA and FDA. Of note, both studies testing trastuzumab plus pertuzumab in the neoadjuvant strategy (i.e. Neosphere [11] and TRYPHAENA [28]) were phase II trials, lacking longer-term efficacy data and unpowered for long-term outcomes like progression-free survival (PFS) and OS. In the adjuvant setting, administration of pertuzumab with trastuzumab showed only a marginal benefit in terms of invasive DFS improvement (0.9%), slightly higher in the high-risk population with node-positive and/or ER-negative HER2⁺ eBC [10]. International guidelines suggest administration trastuzumab-pertuzumab combination along with chemotherapy in high-risk patients, particularly in node-positive patients [29, 30].

Further trials recently focused on additional escalation strategies for high-risk HER⁺ eBC. The KATHERINE study demonstrated a significant benefit from the administration of trastuzumab emtansine (T-DM1) over trastuzumab in patients with residual disease after neoadjuvant therapy, and 1-year of T-DM1 is now the standard of care for these patients [31]. Extended adjuvant anti-HER2 treatment with administration of neratinib after 1 year of trastuzumab also showed a significant recurrence risk reduction, and is currently recommended for patients with hormone receptor (HR)-positive high-risk HER2⁺ eBC [32]. Oppositely, for patients considered at low risk of recurrence, a de-escalation of therapy demonstrated to be non-inferior to standard regimens. An anthracyclines-free chemotherapy with paclitaxel in combination with trastuzumab should then be considered for patients with small (T < 2–3 cm, node-negative), low-risk HER2⁺ eBC [33].

Before the advent of anti-HER2 agents, the OS for metastatic HER2⁺ BC patients was lower than two years [34], with their treatment mostly relying on subsequent lines of traditional chemotherapy. However, the demonstration of the benefit of combining chemotherapy with the anti-HER2 antibody trastuzumab opened a new era in the treatment of this subgroup of patients [34].

Nowadays, most metastatic BC (mBC) patients receive frontline dual blockade with trastuzumab and pertuzumab combined with a taxane, followed by dual blockade maintenance (+/– endocrine treatment in tumors expressing HR) [35]. This regimen has led to an unprecedented OS of 57 months, with more than a third of the patients being alive after 8 years [36]. In case of progression to dual blockade, a standard second-line option is the antibody-drug conjugate (ADC) T-DM1, based on the results of the phase 3 EMILIA study which has proven its superiority to capecitabine plus lapatinib [37]. Nonetheless, this trial did not enroll patients pretreated with pertuzumab, thus activity of TDM-1 in this setting is still to be determined.

In contrast to the first and second line of treatment, no standard treatment is currently recommended for patients progressing to T-DM1 [35]. For about a decade, treatment in this setting has relied on the combination of capecitabine plus lapatinib [38], chemotherapy plus trastuzumab, or the two biological agents (lapatinib + trastuzumab) combined [39]. However, the recent emergence of novel anti-HER2 compounds has redesigned the field, providing highly active alternatives for T-DM1-pretreated patients [40]. More in detail, four of such agents have been recently approved by the FDA in this setting: the ADC trastuzumab deruxtecan, the Fc-engineered anti-HER2 antibody margetuximab and the tyrosine kinase inhibitors (TKIs) tucatinib and neratinib. Trastuzumab deruxtecan and tucatinib are included as treatment options in the last European School of Oncology (ESO)-European Society of Medical Oncologists (ESMO) ABC5 Guidelines [35], based on a favorable magnitude of benefit assessed with the ESMO Magnitude of Clinical Benefit Scale (MCBS) [41]. The small PFS benefit provided by neratinib in the NALA trial [42] was instead not considered enough clinically meaningful to recommend it as a treatment option, nor was the activity of margetuximab, which showed a small PFS benefit in the phase 3 SOPHIA trial [43].

To further improve outcomes of mBC patients, an enlarging arsenal of new drugs targeting HER2 are being developed, encompassing various pharmacological classes (Figure 1). In this section we’ll list some of the most promising new anti-HER2 compounds under development.

BC has been traditionally classified as a poorly immunogenic tumor. Nonetheless, it is a heterogenous disease, and this heterogeneity characterizes also the immunogenic potential of BC. Triple negative and HER2⁺ BC are now recognized as the most immunogenic among BC subtypes [96], with large evidence supporting immunotherapy activity [97]. Indeed, both these subtypes are characterized by a higher programmed cell death ligand-1 (PD-L1) expression [98, 99], tumor mutational burden (TMB) [100, 101] and median percentage of TILs [102, 103], all biomarkers of immune activation.

First clinical studies with immune checkpoint inhibitors (ICIs) in HER2⁺ BC have been conducted in the metastatic setting, with disappointing results.

The JAVELIN trial [104] is a phase I study that tested avelumab monotherapy in patients with mBC, unselected for PD-L1-expression and including all subtypes. Fifteen point five percent of 168 patients enrolled had HER2⁺ BC, but no response was observed among this subgroup of patients.

The phase Ib/II PANACEA study [105] looked instead at the combination of trastuzumab and pembrolizumab in patients with trastuzumab-resistant HER2⁺ mBC. An ORR of 15% was observed in the PD-L1⁺ cohort, whereas there were no responses among PD-L1^– patients. Responses achieved were durable, with a median duration of response of 11.1 months. Of note, median OS was also significantly higher in the PD-L1⁺ cohort than in the PD-L1^– (16.1 vs. 7.0 months; P < 0.01).

Subsequently, other trials were designed to understand if the addition of ICIs could potentially increase durability of approved regimens for HER2⁺ mBC. The phase II KATE2 study investigated the combination of T-DM1 plus atezolizumab or placebo in patients progressing to prior trastuzumab and taxane-based therapy [106]. One hundred thirty three and 69 patients were randomized to receive T-DM1 plus atezolizumab or placebo, respectively, with almost half of the patients pretreated with pertuzumab in both arms. Unfortunately, this trial failed to reach its primary endpoint of PFS benefit. Overall, median PFS was similar in the two treatment arms (8.2 vs. 6.8 months in the atezolizumab and placebo groups, respectively; P = 0.33), as well as response rates. Notably, median PFS was more than doubled in the subgroup of PD-L1⁺ patients (mPFS 8.5 vs. 4.1 months) [106]. In the first-line setting, the addition of atezolizumab to the standard regimen of taxane, trastuzumab and pertuzumab in currently under investigation by the NRG BR004 phase III study (NCT03199885).

Briefly, all trials testing ICIs in HER2⁺ mBC presented so far were formally negative, and failed to provide a benefit in this BC subtype. Nonetheless, biomarker analysis conducted in these trials provided crucial data about the role of TILs and PD-L1 for patients’ selection. In the PANACEA trial, objective responses were observed only in the PD-L1⁺ cohort. Additionally, a preliminary subgroup analysis showed that among PD-L1⁺ tumors, responses were observed almost only in patients with stromal TILs (sTILs) > 5%, whereas few patients with PD-L1⁺/sTILs < 5% responded (ORR 39% vs. 5%) [105]. Similarly, the KATE2 study showed a consistent PFS benefit from the addition of atezolizumab in PD-L1⁺ tumors. Patients whose tumors showed TILs ≥ 5% had also a longer PFS with respect to patients with TILs < 5%, while the opposite was observed in the placebo arm [106].

In the (neo)adjuvant setting, the potential benefit of adding ICIs to standard regimens is also under investigation, with two randomized phase III trial ongoing. The IMpassion 050 study (NCT03726879) is comparing the combination of chemotherapy, trastuzumab and pertuzumab with or without atezolizumab. The APTneo trial (NCT03595592) instead is testing the combination of atezolizumab plus trastuzumab-pertuzumab and two different chemotherapy regimens (with or without anthracyclines), versus the combination of paclitaxel, carboplatin, trastuzumab and pertuzumab. The primary endpoint is pCR for the IMpassion 050, and 5-years event-free survival for the APTneo trial. Both trials are enrolling patients with HER2⁺ eBC unselected for PD-L1-status.

Besides improving outcomes of HER2⁺ BC patients, some of the novel anti-HER2 agents are also revolutionizing the way we interpret HER2 expressions. Indeed, novel anti-HER2 ADCs have recently demonstrated to be active in tumors showing low HER2 expression [immunohistochemical (IHC) staining 1+ or 2+ with negative in situ hybridization assay], which account for about half of all BCs [107]. These tumors are not expected to be dependent on HER2 signaling pathway, and indeed HER2 blockade with trastuzumab or pertuzumab has shown to be poorly active in this setting [108, 109].

The most relevant activity to date in HER2-low mBC was observed with trastuzumab deruxtecan, which achieved a response rate of 37% and a PFS of 11 months in highly pretreated (median of 7.5 prior therapies) patients [110]. Interestingly, the drug showed a similar activity in tumors with a HER2 IHC score of 1+ and 2+, whereas ORR numerically differed based on HR expression, with luminal-like tumors showing the highest response rate. Data of trastuzumab deruxtecan in combination with nivolumab were also recently reported at San Antonio [111]. Overall, 16 highly pretreated HER2-low mBC patients were treated with the combo, achieving an ORR of 38% and a mPFS of 6.3 months. Of note, toxicity of the regimen appeared comparable with that of trastuzumab deruxtecan monotherapy, with 10% of the patients experiencing ILD. The randomized phase 3 DESTINY-Breast04 and DESTINY-Breast06 trials are currently comparing trastuzumab deruxtecan to chemotherapy in pretreated HER2-low mBC patients, and may lead to the confirmation of the HER2-low paradigm.

Trastuzumab duocarmazine was also tested in a cohort of highly pretreated HER2-low mBC patients, showing an ORR of 28–40% and a median PFS of 4–5 months (depending HR expression). However, no confirmatory trial was yet initiated with the compound in HER2-low mBC.

Preclinical observations suggest that the activity observed with the abovementioned ADCs in HER2-low cancers may be dependent on the delivery of cytotoxic payload, rather than the inhibition of the HER2 pathway [46, 112]. In this sense, HER2 may be considered a “targetable” biomarker in HER2-low cancers, whereas it could be also considered an “actionable” pathway in HER2⁺ cancers. If ongoing randomized trials confirmed this paradigm, we may expect anti-HER2 ADCs to become useful treatment options for up to 70% of all mBC patients, with an important redefinition of current BC treatment algorithms [107].

Despite impressively improving outcomes of BC patients, anti-HER2 treatments are still subject to resistance generation through several mechanisms, both in the early and advanced settings. Such mechanisms have been extensively studied, although mostly preclinically and in the context of single HER2-blockade.

HER2⁺ cancer cells can develop resistance to HER2-targeting agents by impairing drug binding to the HER2 receptor, by constitutively activating parallel or downstream signaling pathways, trough metabolic reprogramming or reduced immune system activation [113]. The first mechanism, namely impairing the adequate binding of the compound, appears particularly relevant in the context of treatment with ADCs [114]. Indeed, by lowering HER2 expression, impairing the internalization of the conjugate or the intracellular release and permanence of the payload, cancer cells may develop resistance to HER2-targeting ADCs. Some clinical confirmations of the importance of HER2 expression level come from clinical trials testing T-DM1, where higher HER2 protein of mRNA expression led to higher activity of the compound [115, 116], whereas clinical confirmations are pending for the remaining mechanisms.

Further work is required to fully understand and overcome mechanisms of resistance to traditional anti-HER2 agents, as well as to investigate resistance mechanisms to the novel anti-HER2 agents which have recently entered the clinic.

In about twenty years since the approval of the first anti-HER2 agents, a multiplicity of other anti-HER2 agents have been approved, significantly improving our ability to cure BC patients in the early setting, and greatly extending their survival in the advanced setting. Recent evolutions in this field promise to improve outcomes even further, through the engineering of novel, highly active anti-HER2 TKIs and ADCs, as well as the exploration of new ways of targeting HER2, such as bsAbs, vaccines and adoptive cell therapies. Importantly, novel agents harbor the potential to extend the benefit of anti-HER2 treatments to the wide cohort of HER2-low BC patients, a population for which no anti-HER2 agent is yet approved. Large ongoing randomized trials will determine if these promises are kept, and may relevantly reshape BC treatment algorithms in the years to come.