Intrinsic and Acquired Resistance to HER2-Targeted Therapies in HER2 Gene-Amplified Breast Cancer: Mechanisms and Clinical Implications

Abstract

Approximately 25% of human breast cancers overexpress the HER2 (ErbB2) proto-oncogene, which confers a more aggressive tumor phenotype and associates with a poor prognosis in patients with this disease. Two approved therapies targeting HER2, the monoclonal antibody trastuzumab and the tyrosine kinase inhibitor lapatinib, are clinically active against this type of breast cancer. However, a significant fraction of patients with HER2+ breast cancer treated with these agents eventually relapse or develop progressive disease. This suggests that tumors acquire or possess intrinsic mechanisms of resistance that allow escape from HER2 inhibition. This review focuses on mechanisms of intrinsic and/or acquired resistance to HER2-targeted therapies that have been identified in preclinical and clinical studies. These mechanisms involve alterations to HER2 itself, coexpression or acquisition of bypass signaling through other receptor or intracellular signaling pathways, defects in mechanisms of cell cycle regulation or apoptosis, and host factors that may modulate drug response. Emerging clinical evidence already suggests that combinations of therapies targeting HER2 as well as these resistance pathways will be effective in overcoming or preventing resistance.

I. INTRODUCTION

HER2 is a member of the ErbB family of receptor tyrosine kinases (RTKs), which includes EGFR (ErbB1), HER3 (ErbB3), and HER4 (ErbB4). Binding of ligands to the extracellular domain of EGFR, HER3, or HER4 (HER2 has no known ligand) induces homo- and heterodimerization of receptors followed by phosphorylation of several tyrosines in the receptors’ C-termini. In turn, these residues serve as docking sites for a number of src homology (SH)2-containing adaptors and signal transducers that mediate the transforming effects of this receptor network.¹ Of the ErbB family members, HER2 is the preferred dimerization partner and possesses the most potent kinase catalytic activity capable of amplifying signaling through ErbB co-receptors. Heterodimers of HER2 with kinase-inactive HER3 are the most transforming of the ErbB network. They potently activate the phosphatidylintositol 3-kinase (PI3K)-Akt survival pathway via HER2-mediated phosphorylation of HER3, which, in turn, recruits the p85 regulatory subunit of PI3K.^2–6

Amplification of HER2 occurs in approximately 25% of human breast cancers. Prior to the development of HER2-targeted therapies, HER2 over-expression was associated with poor patient outcome.^7,8 There are two FDA-approved therapies that target HER2 in breast cancer: (1) trastuzumab, a monoclonal antibody directed against the extracellular domain of HER2, and (2) lapatinib, a small molecule ATP-competitive reversible kinase inhibitor of HER2 and EGFR. These therapies have shown significant clinical benefit, including a 50% reduction in the risk of death after treatment with trastuzumab or a similar improvement in disease-free survival with lapatinib after progression on trastuzumab.^9–13 However, most patients with advanced HER2 gene-amplified breast cancer eventually relapse after treatment, suggesting that tumors acquire or intrinsically possess mechanisms for escape from HER2 inhibition. Clinical efficacy of HER2-directed therapies appears to be limited to breast cancers that overexpress the HER2 protein as determined by immunohistochemistry (IHC), or exhibit gene amplification of HER2 as determined by fluorescence in situ hybridization (FISH).¹⁴ Thus, HER2 overexpression as defined by these two assays is currently the only biomarker predictive of response to HER2-targeted therapy.

Trastuzumab is a humanized IgG₁ that recognizes an epitope in the extracellular domain of HER2. The therapeutic effect of the antibody is likely due to a number of effects, including antibody-mediated immunomodulatory activities as well as disruption of HER2 function and downstream signaling. Both the innate and adaptive immune responses are required for the therapeutic effect of trastuzumab when tested in mouse models.^15,16 Trastuzumab also inhibits tumor cell growth via induction of G1 cell cycle arrest associated with induction of the CDK inhibitor p27^KIP1.^17–20 Binding of trastuzumab to the juxtamembrane region of HER2 disrupts ligand-independent HER2-HER3 heterodimerization.²¹ This results in partial inhibition of PI3K-Akt signaling, a major effector of HER2-mediated tumor progression.²²

Lapatinib is an ATP-competitive reversible tyrosine kinase inhibitor (TKI) of HER2 and EGFR. In HER2+ breast cancer, lapatinib blocks HER2 phosphorylation, resulting in inhibition of PI3K-Akt and MAPK downstream of the HER2 receptor. Blockade of these pathways inhibits cell viability in vitro and xenograft growth in vivo.^23,24 In patients with metastatic HER2+ breast cancer, treatment with lapatinib induces variable levels of signaling inhibition and apoptosis.²⁵ In a recently reported neoadjuvant trial in patients with newly diagnosed HER2+ breast tumors, the major effect of single-agent lapatinib was inhibition of cell proliferation without a detectable induction of tumor cell apoptosis.²⁶ Conversely, treatment with trastuzumab, which has less profound effects on signaling in vitro than does lapatinib, was shown to induce apoptosis in primary HER2+ tumors,²² suggesting the antibody effect may involve a host-antibody interaction not detectable in tumor cell autonomous experiments in vitro.

In patients with metastatic disease, the therapeutic action of trastuzumab and lapatinib tends to be short lived. Even though some patients treated in the adjuvant setting will likely be cured of their cancer, it is anticipated that a fraction will eventually recur. This suggests that tumors harbor de novo or acquired mechanisms of resistance to therapeutic inhibitors of HER2. We will discuss potential mechanisms of drug resistance that have been identified in preclinical models and, in some cases, confirmed in patient tumors. These occur at three levels: first, mechanisms intrinsic to the target itself; second, resistance that involves parallel bypass signaling pathways to overcome HER2 inhibition; and third, resistance that arises from defects in the apoptosis pathway in tumor cells or in extrinsic host factors that participate in drug action (Fig. 1).

Figure 1.

Mechanisms of resistance to HER2 inhibitors. A. Resistance mechanisms involving alterations in HER2 that abrogate trastuzumab binding include expression of alternative isoforms of HER2 (p95, Δ16) or expression of mucins that block the binding of trastuzumab. B. Increase in ligand production enables signaling through other ErbB family members to overcome inhibition of HER2. C. Mutations that activate intracellular signaling pathways downstream of HER2 and uncouple signaling from HER2 inhibition include mutations in PIK3CA and/or loss of function of PTEN. D. Signaling through bypass pathways (e.g., emanating from MET, IGF-1R, EphA2, or EpoR) engage the downstream effectors of HER2 such as PI3K-Akt either directly or through intracellular kinases such as Src.

II. INTRINSIC HER2 ALTERATIONS

A mechanism of resistance to inhibitors of HER2 is mutation of the target itself, resulting in alteration of drug binding. This mechanism is exemplified by the acquired so-called “gatekeeper” kinase domain mutations observed in patients with lung cancer treated with EGFR TKIs and patients with CML and gastrointestinal stromal tumors treated with imatinib.^27–32 For HER2, this would also include mutations in the juxtamembrane region that contains the binding epitope of trastuzumab. Anido et al. described p95-HER2, a truncated form of HER2 lacking the antibody binding region, which arises from alternate transcription initiation sites in HER2.³³ This form of HER2 retains kinase activity and is susceptible to inhibition by lapatinib but not trastuzumab.³⁴ Patients with metastatic breast cancer harboring cytosolic expression of p95-HER2 exhibit a very low response rate to treatment with trastuzumab and chemotherapy compared to those patients without p95-HER2 in their tumors. Conversely, tumors with p95-HER2 are still susceptible to kinase inhibition with a TKI, as was suggested by a similar response rate to capecitabine and lapatinib observed in patients with breast cancer with and without p95-HER2.³⁵ A recent study reported a nuclear localized truncated form of HER2, also 95 kDa in size, which retains phosphorylation and nuclear localization upon treatment with lapatinib.³⁶ The frequency and clinical significance of this finding are unknown at this time.

A splice variant that eliminates exon 16 in the extracellular domain of the HER2 receptor has also been identified in HER2+ breast cancers and cell lines.^37,38 Cell lines expressing this D16 HER2 isoform are resistant to trastuzumab.^38,39 This variant does not eliminate the trastuzumab epitope on HER2, but does appear to stabilize HER2 homodimers and may potentially prevent their disruption upon binding by the antibody.³⁸ In addition, the D16 isoform was found to interact directly with the Src tyrosine kinase, and treatment with the Src inhibitor dasatinib overcame the resistance to trastuzumab conferred by the alternative splicing variant.³⁹ A role for Src kinases in HER2 inhibitor resistance will be further discussed below.

Point mutations or small insertions in the HER2 gene have been identified in other cancers. A small number (2%–4%) of non-small-cell lung cancers (NSCLC), as well as gastric, colorectal, and head and neck cancers, have been found to have alterations in the HER2 gene.^40–46 These include primarily amino acid substitutions or insertions localized in the kinase domain. An insertion in exon 20, originally identified in NSCLC, was able to confer resistance to lapatinib and trastuzumab when expressed in breast cancer cell lines.⁴⁷ HER2 mutations have been reported in a small number of human breast cancers but in the absence of HER2 gene amplification.⁴¹ To our knowledge, HER2 mutations in HER2-overexpressing breast tumors have not been reported to date. One possible reason is that these mutations may comprise only a portion of the amplified HER2 alleles and, therefore, exist below the limits of sensitivity of traditional DNA sequencing methods. Nonetheless, using a next-generation sequencing approach with higher sensitivity that was measured to detect a variant frequency as low as 0.08%, no mutations in HER2+ breast cell lines or tumors were detected.⁴⁸ Another possibility is that these mutations may be selected for or acquired only after the selective pressure of anti-HER2 treatment. If so, they are likely to be detected in tumors that are progressing after primary HER2-targeted therapy.

Finally, unique for the case of trastuzumab is the possibility of abrogating drug binding to the target by coexpression of another protein that binds to the drug target. For example, mucin-4 (MUC4), a membrane-associated glycoprotein, when overexpressed can co-localize with HER2 and mask the binding site for trastuzumab.^49,50 A cleaved form of another mucin family member, MUC1*, was also found to be overexpressed in a cell line selected for trastuzumab resistance.⁵¹ While this cleaved MUC1 isoform has been shown to associate with HER2, and the interaction is enhanced by ligands,⁵² other studies have shown that MUC1* can homodimerize and induce signaling for survival and proliferation on its own.^53,54 Figure 1A illustrates several of these alterations in HER2 that are thought to play a role in resistance.

III. ACTIVATION OF COMPENSATORY SIGNAL TRANSDUCTION PATHWAYS

A. Receptor Tyrosine Kinases (RTKs)

Signaling through other RTKs can transactivate HER2 and amplify signal transduction downstream, thus bypassing the inhibitory effect of lapatinib or trastuzumab. HER2 is known to heterodimerize with other ErbB family members, and signaling initiated by ligands of EGFR, HER3, or HER4 can rescue the antiproliferative effects of trastuzumab.^55,56 In a model of acquired resistance to trastuzumab in breast cancer xenografts, our laboratory reported increased expression of EGFR and HER3 ligands, resulting in activation of EGFR and HER3 as well as increased EGFR/HER2 heterodimers in trastuzumab-resistant cells (Fig. 1B).⁴ This is consistent with data showing that trastuzumab is unable to block ligand-induced heterodimerization of HER2.⁵⁷ Another mechanism for increased availability of ErbB ligands associated with trastuzumab resistance is activation of TGFb receptors that, in turn, activate the sheddase TACE/ADAM17 to release more TGFa, amphiregulin, and heregulin. This increase in ligand activates HER3. A gene signature of TGFb activity, derived from gene expression profiling after expression of a constitutively active mutant of the TGFb type I receptor (ALK5), correlates with resistance to trastuzumab and poor clinical outcome in patients.⁵⁸ Of note, these models of trastuzumab resistance appear to remain dependent on HER2 signaling, as treatment with lapatinib to block HER2 kinase activity directly, or with pertuzumab, an antibody that blocks HER2 heterodimerization, restores trastuzumab action.^4,57,59,60

Cross-talk with receptors outside of the ErbB family has also been implicated as a means to bypass trastuzumab action (Fig. 1D). Overexpression of the IGF-1 receptor or an increase in levels of IGF-1R/HER2 heterodimers can potently activate PI3K-Akt signaling and confer resistance to trastuzumab.^61,62 IGF-1R has been found to complex with both HER2 and HER3 in a heterotrimeric complex.⁶³ Inhibition of IGF-1R with a neutralizing antibody or a small molecule TKI, or targeting the HER2 kinase with lapatinib, was found to overcome IGF-1R-mediated resistance to trastuzumab.^62,64,65 Recently, targeting the IGF-1R axis with metformin was also shown to overcome resistance to the antibody.⁶⁶ In a neo-adjuvant trial of chemotherapy plus trastuzumab, a high level of IGF-1R expression measured by IHC correlated with a poor clinical response.⁶⁷ In another study, however, IGF-1R levels in pre-treatment biopsies did not correlate with response to trastuzumab or to clinical outcome.⁶⁸ One explanation is that the IGF-1R expression is selected for after acquired resistance and, therefore, the pre-treatment levels of IGF-1R are not predictive. Indeed, in some studies, increased IGF-1R or IGF-1R/HER2 complexes were observed only after development of acquired resistance to trastuzumab.^62,63,66

The RTK MET has also been implicated in trastuzumab resistance. MET has been shown to interact with HER3 to engage PI3K signaling downstream of EGFR in gefitinib-resistant NSCLC.⁶⁹ Upon treatment with trastuzumab, HER2+ breast cancers up-regulate MET levels. HGF-induced signaling through MET has been shown to abrogate the action of trastuzumab. Further, overexpression of MET and its ligand, HGF, were reported in a cohort of HER2+ patients who did not respond to chemotherapy plus trastuzumab.⁷⁰

Overexpression of the EphA2 receptor was found to predict reduced disease-free and overall survival in a cohort of patients with HER2+ breast cancer. When overexpressed in cell lines, EphA2 conferred resistance to trastuzumab. Moreover, cell lines with acquired trastuzumab resistance showed increased activation of EphA2, and treatment with an EphA2 neutralizing antibody restored trastuzumab sensitivity.⁷¹ Most recently, the erythropoietin receptor was found to be co-expressed in cell lines and primary tumors that overexpress HER2. In these cell lines, concurrent treatment with recombinant human erythropoietin conferred trastuzumab resistance. This was mediated through signaling through Jak and Src, leading to inactivation of the lipid phosphatase PTEN. Finally, in patients with HER2+ metastatic breast cancer, the concurrent administration of eryth-ropoietin and trastuzumab correlated with a shorter progression-free and overall survival compared to patients not receiving erythropoietin.⁷²

Several groups have shown an increase in HER3 transcription and HER3 phosphorylation after short-term treatment with lapatinib.^73–75 This response follows the inhibition of AKT and AKT-mediated phosphorylation of FoxO followed by translocation of FoxO to the nucleus and derepression of FoxO-mediated HER3 transcription. Although not necessarily a mechanism of resistance, it is of interest that inhibition of HER3 with genetic or pharmacological means greatly enhances the action of lapatinib. Finally, in a model of acquired resistance to lapatinib, the RTK AXL was up-regulated in the resistant cells. AXL potently engaged p85 to activate PI3K and bypass the effects of either lapatinib or trastuzumab. A multikinase inhibitor with activity against AXL restored sensitivity to HER2 antagonists.⁷⁶

B. Intracellular Kinases

Mutations in components of the PI3K/AKT pathway are the most frequent tumor somatic alterations in breast cancer, altogether occurring in >30% of invasive tumors. These include gain of function mutations in PIK3CA, the gene encoding the p110a catalytic subunit of PI3K, AKT1, or PIK3R1 (p85a), amplification of PIK3CA or AKT2, loss of PTEN, the lipid phosphatase that dephosphorylates PIP3 and negatively regulates PI3K, and loss of the tumor suppressor INPP4B.^77–83 It is generally accepted that the antitumor activity of HER2 inhibitors depends on inhibition of PI3K-AKT downstream of HER2.^6,21 Thus, it is not surprising that activating mutations in the PI3K pathway itself can confer resistance to HER2 inhibitors.

Using a large-scale siRNA genetic screen, Berns et al. identified PTEN as the only gene whose knockdown conferred trastuzumab resistance. Further, induced overexpression of gain-of-function PIK3CA mutants similarly conferred trastuzumab resistance. Finally, patients with “hot spot” PIK3CA mutations and undetectable or low PTEN measured by IHC, exhibited a poorer outcome after treatment with chemotherapy and trastuzumab compared to patients without those alterations.⁸⁴ An earlier study had also shown that loss of PTEN correlates with a lower response to trastuzumab in patients.⁸⁵ More recently, Esteva et al. found that PI3K pathway activation, defined as PTEN loss and/or PIK3CA mutation, was associated with a poor response to trastuzumab as well as a poorer overall survival.⁸⁶ In addition, human breast cancer cell lines containing endogenous mutations in PIK3CA are intrinsically resistant to trastuzumab.^6,21,87,88 Human breast cancer cell lines in which PIK3CA mutations are ectopically expressed exhibit an attenuated response to lapatinib.⁸⁹ On the other hand, loss of PTEN has not been consistently associated with resistance to lapatinib.⁹⁰ The centrality of the PI3K pathway in de novo and acquired drug resistance is further underscored by the effects of inhibitors of PI3K/AKT in resistant cells. For example, combinations of trastuzumab with the PI3K inhibitor XL147, or trastuzumab or lapatinib with the dual PI3K-mTOR inhibitor BEZ235, inhibit growth of PIK3CA mutant xenografts resistant to anti-HER2 therapies.^89,91,92 In a model of trastuzumab resistance caused by PTEN loss, targeting mTOR or AKT was able to at least partially overcome resistance.⁹³ Evidence is also emerging from the clinic that targeting the PI3K axis in addition to HER2 may be a strategy to overcome resistance. For example, in a phase II study, the combination of the TORC1 inhibitor everolimus with trastuzumab and chemotherapy showed a partial response rate of 19% and a clinical benefit rate of 81% in patients with HER2+ metastatic breast cancer that had previously shown progression on trastuzumab plus taxanes.⁹⁴ Some of these alterations in the PI3K/ATK pathway are diagrammed in Fig. 1C.

Using a mass spectrometry approach to identify aberrant tyrosine phosphorylation in HER2+ breast cancer cells with acquired resistance to lapatinib, we found up-regulation of Src family kinase (SFK) activity, particularly Yes, in several of the resistant cell lines. Resistance was associated with recovery of PI3K-AKT signaling despite continued inhibition of HER2 with lapatinib. Combination of lapatinib and a SFK inhibitor partially blocked PI3K activation in the resistant cells and restored sensitivity to lapatinib in BT474 xenografts.⁹⁵ There also appears to be a role for Src activity in resistance to trastuzumab. In trastuzumab-resistant HER2+ cells, PTEN was no longer capable of dephosphorylating Src. Addition of Src inhibitor to trastuzumab overcame trastuzumab resistance.⁹⁶ Src activity is also involved in the resistance conferred by the D16 HER2 isoform, where combination of trastuzumab with Src inhibitor was able to restore trastuzumab action.³⁹ Src kinase mediates at least part of the resistance conferred by activation of the erythropoietin receptor. EpoR activates Src via Jak2 and Src associates with HER2, where it is proposed to phosphorylate and inhibit PTEN. The inhibition of PTEN, in turn, up-regulates PI3K, providing a viable mechanism for attenuation of trastuzumab action.⁷² Finally, the activation of EphA2 seen after chronic trastuzumab treatment and onset of trastuzumab resistance is also mediated through Src.⁷¹

IV. DEFECTS IN APOPTOSIS AND CELL CYCLE CONTROL

In “oncogene-addicted” tumors, such as HER2+ breast cancer, intracellular survival and proliferation signals are under the dominant control of a single oncogene. In these tumors, targeting that oncogene product is expected to not only arrest proliferation but also induce apoptosis. Therefore, it is not surprising that alterations in the normal apoptotic machinery are emerging as important causes of resistance to oncogene-targeted therapies. In a recent report, levels of the pro-apoptotic BH3-only Bcl2 family member BIM were shown to be predictive of response to targeted therapy, not only in HER2+ breast cancer, but also in NSCLC harboring EGFR mutations and other cancers with mutations in the PIK3CA gene. In this study, although lapatinib inhibited HER2 and downstream signaling in all HER2+ breast cancer cell lines, only the cell lines with high basal levels of BIM underwent apoptosis.⁹⁷ This suggests that BIM levels might be a biomarker predictive of a likelihood of response to a TKI in an oncogene-addicted cancer. Another group reported similar results in HER2 gene-amplified breast cell lines with and without activating mutations. Lapatinib-induced apoptosis was associated with BIM up-regulation and depended on basal levels of BIM expression. In PIK3CA mutant cells, however, both the growth inhibitory effect and induction of BIM following treatment with lapatinib were blunted.⁹⁸

Survivin is a member of the inhibitor of apoptosis (IAP) protein family, which inhibits the activity of caspases, important effectors of programmed cell death. Modulation of survivin levels has been shown to be a point of convergence of several pathways that bypass the action of HER2 inhibitors. HER2+ breast cancer cells with acquired resistance to lapatinib up-regulated ERa, which, in turn, induced FoxO3a-mediated transcription of survivin. High levels of survivin allowed for escape of these cells from lapatinib.⁹⁹ Elevated levels of survivin and the antiapoptotic Mcl-1 proteins were found in trastuzumab-resistant cells; treatment with a broad-spectrum kinase inhibitor that reduces levels of survivin and Mcl-1 inhibited the growth and survival of the cells.¹⁰⁰ Knockdown of survivin followed by trastuzumab restored sensitivity to the HER2 antibody.¹⁰¹ Furthermore, survivin expression is regulated by PI3K signaling in HER2-amplified breast cancer cells. Inhibition of HER2-PI3K reduces survivin expression simultaneous with drug-induced apoptosis in HER2-overexpressing breast cancer cells.^98,102

Altered control of progression through the cell cycle in response to HER2 inhibition also plays a role in resistance. Cell lines made resistant to trastuzumab by chronic exposure showed focal amplification of cyclin E. In a cohort of patients with HER2+ breast cancers treated with trastuzumab, amplification of cyclin E was associated with lower response to trastuzumab. Finally, CDK2 inhibitors reduced growth of trastuzumab-resistant xenografts.¹⁰³ Alternatively, down-regulation of the Cdk inhibitor p27^KIP1 and a resulting increase in Cdk activity is also associated with trastuzumab resistance.¹⁰⁴ Modulation of levels of p27^KIP1 appears to be a common endpoint for several of the resistance pathways noted above, including signaling from IGF-1R and MET.^61,62,70 These data are in agreement with earlier results in which treatment with trastuzumab in antibody-sensitive cells resulted in an increase in p27 levels and nuclear localization.⁶

V. HOST FACTORS REQUIRED FOR TRASTUZUMAB ACTION

The resistance mechanisms to HER2 inhibition described thus far have been tumor cell autonomous. However, there is strong evidence to support that trastuzumab exerts its antitumor effect via the engagement of host immune effectors and subsequent antibody-dependent, cell-mediated cytotoxicity (ADCC) in addition to HER2 down-regulation and inhibition of post-receptor signaling. Thus, host factors that affect this immunomodulatory function can also contribute to trastuzumab resistance. In mice lacking FcgRIII and, thus, deficient in NK cells and macrophages capable of binding the Fc region of trastuzumab, the therapeutic effect of trastuzumab was markedly diminished.¹⁶ Consistent with this important preclinical experiment, polymorphisms in the gene encoding FcgRIII in humans are associated with response to trastuzumab. Notably, PBMCs from patients with polymorphisms associated with an improved outcome after trastuzumab induce a stronger trastuzumab-mediated ADCC in vitro.¹⁰⁵ Moreover, an inability of host (patient) cells to mediate a sufficiently strong ADCC response may contribute to de novo resistance, as leukocytes from those patients who did derive benefit from trastuzumab show a higher level of in vitro ADCC activity than those who did not.¹⁰⁶ A follow-up study found that the quantity and lytic efficiency of CD16(+) lymphocytes are the major factors to affect the level of ADCC induction by trastuzumab, which, in turn, correlated with tumor response.¹⁰⁷

VI. OVERCOMING RESISTANCE WITH COMBINATIONS OF ANTI-HER2 THERAPIES

The majority of patients with HER2+ metastatic breast cancer progress after an initial clinical response of variable duration. Several lines of clinical data suggest that HER2+ breast cancers continue to depend on HER2 at the time of progression. For example, the HER2 TKI lapatinib is effective in patients who progress on trastuzumab.⁹ Furthermore, continuation of trastuzumab in combination with chemotherapy after progression on trastuzumab is associated with a better progression-free survival compared to chemotherapy alone.¹⁰⁸ A post-hoc analysis of this last study showed that patients receiving anti-HER2 treatment even as third-line therapy had a superior progression-free survival compared to those who did not continue HER2-directed therapy.¹⁰⁹ Like lapatinib, the dual EGFR/HER2 TKI neratinib has shown clinical activity in patients with HER2+ metastatic breast cancer who progress on trastuzumab.¹¹⁰ Moreover, the combination of lapatinib plus trastuzumab was more effective than lapatinib alone following progression on trastuzumab.¹¹¹ More recently, the NeoALTTO study compared trastuzumab, lapatinib, and both inhibitors, in combination with paclitaxel for 12 weeks prior to breast surgery. The combination arm exhibited a higher rate of pathologic complete response (51.3%) than either the trastuzumab (29.5%) or lapatinib (24.7%) arms.¹¹²

Other combination strategies that exploit this continued dependence on HER2 at the time of acquired resistance are in clinical development. The antibody-toxin fusion trastuzumab-DM1 consists of trastuzumab conjugated via a non-cleavable linker to a derivative of maytansine, an inhibitor of microtubule polymerization. T-DM1 has shown high tolerability and remarkable clinical activity in phase I–II trials. T-DM1 induces a 25% clinical response rate in heavily pretreated patients with HER2-overexpressing metastatic breast cancer who had progressed on trastuzumab, lapatinib, taxanes, and anthracyclines.^113,114 Even though it is used at lower doses and frequency than trastuzumab, T-DM1 can still inhibit signaling and engage an ADCC response in lapatinib-resistant xenografts.¹¹⁵ Targeting HER2 with trastuzumab in combination with pertuzumab, an antibody that binds to an epitope in the extracellular heterodimerization domain of HER2, is also an effective strategy. Treatment with pertuzumab in combination with trastuzumab, but not either alone, was effective in trastuzumab-resistant xenografts and in patients who had previously progressed on trastuzumab.^116,117 It is proposed that both antibodies are required to completely disable ligand-induced and ligand-independent HER2-HER3 heterodimers and, therefore, maximally inhibit PI3K signaling.²¹ In the recent neoadjuvant NeoSphere trial, patients with newly diagnosed HER2+ breast cancer treated with chemotherapy plus trastuzumab and pertuzumab in combination achieved a pathologic complete response rate of 45.8%, significantly higher than the 29% rate in the chemotherapy plus trastuzumab arm.¹¹⁸ In Table 1, we summarize several combinations of anti-HER2 targeted therapies and their rationale.

TABLE 1.

Two Drug Combinations of Anti-HER2 Therapies and Mechanisms of Action

Combination	Mechanism(s) of action
trastuzumab + lapatinib (or neratinib, afatinib)	ADCC, partial disruption of HER2-HER3 dimers, inhibition of HER2/EGFR kinases
trastuzumab + pertuzumab	Disruption of ligand-dependent and -independent HER2-HER3 dimers, ADCC
T-DM1 + pertuzumab	ADCC, disruption of ligand-dependent and -independent HER2-HER3 dimers, inhibition of microtubules with targeted chemotherapy (DM1)
trastuzumab + PI3K/AKT pathway inhibitor	ADCC, partial disruption of HER2-HER3 dimers, inhibition of PI3K/AKT
T-DM1 + PI3K/AKT pathway inhibitor	ADCC, partial disruption of HER2-HER3 dimers, inhibition of PI3K/AKT, inhibition of microtubules with targeted chemotherapy (DM1)
Lapatinib (or neratinib, afatinib) + PI3K/AKT pathway inhibitor	Simultaneous inhibition of HER2/EGFR kinases and PI3K/AKT
trastuzumab + Src inhibitor	ADCC, partial disruption of HER2-HER3 dimers, inhibition of Src tyrosine kinase
trastuzumab + TORC1 inhibitor (i.e., everolimus)	ADCC, partial disruption of HER2-HER3 dimers, inhibition of TORC1 kinase
trastuzumab or lapatinib + HER3 neutralizing antibody	Adds blockade of HER3 ligand binding and/or HER3 receptor down-regulation to trastuzumab or lapatinib

VII. CONCLUSIONS

Several mechanisms of resistance to both trastuzumab and lapatinib have been identified in preclinical studies. Few of these have been prospectively validated in the clinic. However, there is enough evidence to suggest that some of them do limit the effectiveness of HER2-directed therapy, particularly when these therapies are used as single agents. Emerging clinical data suggest the concept that combinations of therapies targeting the HER2 signaling network at multiple points early in the natural history of HER2+ breast cancer can abrogate drug resistance. What remains a challenge is determining the resistance mechanism(s) in a particular patient so that potential therapies can be individualized. This will require commitment to rebiopsy and molecular analysis of recurrent tumors after progression on primary HER2-targeted therapy. Alternatively, the increasing use of pre-operative therapy should provide a clinical research platform where combinations of anti-HER2 agents can be compared using pathological complete response in the breast as a clinical endpoint predictive of long-term outcome. Another benefit of a pre-operative platform is that residual tumor tissue is available at the time of surgery. These “drug resistant” residual cancers may well reflect the molecular profile of drug-resistant micrometastases and can be interrogated with open-ended molecular approaches to identify biomarkers and/or effectors of resistance to anti-HER2 therapies.

Abbreviations

RTK

receptor tyrosine kinase

SH2

Src homology 2

PI3K

phosphatidylinositol 3-kinase

FDA

Food and Drug Administration

ATP

adenosine triphosphate

IHC

immunohistochemistry

FISH

fluorescence in situ hybridization

CDK

cyclin dependent kinase

TKI

tyrosine kinase inhibitor

kDa

kilodalton

NSCLC

non-small-cell lung cancer

PIP3

phosphatidylinositol (3,4,5)-trisphosphate

SFK

Src family kinase

IAP

inhibitor of apoptosis

ADCC

antibody-dependent cell-mediated cytotoxicity

PBMC

peripheral blood mononuclear cells

T-DM1

trastzumab emtansine (trasutuzumab conjugated derivative of maytansine 1)