Reviews in Oncology Reviews in Oncology CIC Edizioni Internationali 2013 January-March; 1(1): 33–40. ISSN: 2282-6378

When HER2 inhibition fails: hindering compensatory mechanisms through combinatorial approaches

Danila Serpico and Serena Di Cosimo

Department of Medical Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Approximately 20% of invasive breast cancer (BC) cases exhibit gene amplification or protein overexpression of the Human Epidermal growth factor Receptor 2 (HER2). As compared to the rest of tumors, those with HER2 amplification and/or over expression (hereinafter HER2 positive BC) show poor prognosis and a more aggressive phenotype with early relapses and increased rate of breast specific death. A decade ago, phase III studies revealed that combining the humanized monoclonal antibody anti-HER2 trastuzumab with chemotherapy increased response rates, time to disease progression, and overall survival as compared with chemotherapy (1, 2) or trastuzumab alone in patients with HER2 positive metastastic BC (3). More recently, trastuzumab, in combination or after adjuvant chemotherapy, has been demonstrated to improve disease-free and overall survival rates in patients with early stage breast cancer as well as pathologic complete remission rate, disease free survival and overall survival in the neoadjuvant setting (47).

Therefore, trastuzumab has become a mainstay for the treatment of women diagnosed with HER2 positive breast cancer and is now approved for the treatment of HER2-overexpressing metastatic breast cancer in combination with a taxane for first line therapy or as single agent in patients who have received one or more chemotherapy regimens for metastatic disease; trastuzumab is also indicated for adjuvant treatment of HER2 overexpressing node positive or node negative/high risk breast cancer as part of a treatment regimen consisting of doxorubicin, cyclophosphamide, and a taxane, or with docetaxel and carboplatin or as a single agent following multi-modality anthracycline based therapy.

However, most of the patients with HER2 advanced disease progress after a variable period of time. At that point, resistance to therapy is not only common but expected. In addition, 15% of patients treated with adjuvant trastuzumab containing regimens inevitably relapse. Herein we summarize the general mechanisms of resistance to trastuzumab and we report an overview of novel anti-HER2 targeted therapies and strategies.

This systematic review focused on the mechanisms of resistance to HER2-targeted therapies. A computerised literature search, as of the 1st December 2012, was conducted by the Authors using PubMed with no language restrictions. The following search terms were used: “trastuzumab resistance”, “lapatinib resistance”, “pertuzumab resistance”, “anti-HER2 targeted-therapy resistance”, “novel drug HER2 pathway”. Two independent Authors reviewed the results of the searches and the studies to be included. The references of all eligible full-text articles were examined for potentially relevant studies.

In patients with HER2 positive metastatic breast cancer, response to trastuzumab as single agent ranges from 34–48% in the first line, with a time to disease progression of 3.5–3.8 months (2 mg/kg and 4 mg/kg weekly, respectively) and a median duration of survival of 22.9–25.8 months, to 18% in the second and third line setting (8, 9). The main mechanisms of resistance to trastuzumab include: alterations in extracellular domain of HER2; over-expression of ligand; interactions with other receptors; activation of downstream signaling or alternative and compensatory pathways (Fig. 1, Tab. 1).

Table 1 Table 1
Mechanisms of resistance to trastuzumab.

Alterations in binding site
Resistance to targeted antibodies may develop through the blockade of the interaction between the antibody and its targeted protein. The binding of trastuzumab with HER2 induces a complex that is then internalized and degraded by the proteasome. Preclinical models show that resistance to trastuzumab may be due to increased expression of proteins that sterically hinder HER2 from binding to trastuzumab. Among such proteins, there is the membrane-associated glycoprotein MUC4, which activates the receptor through its epidermal growth factor (EGF)-like domain on the ASGP-2 subunit. Through this interaction, it is proposed that MUC4 serves as a ligand for HER2, resulting in increased phosphorylation of HER2 on the residue Tyr 1248, a major phosphorylation site contributing to the transforming ability of HER2. In addition, elevated MUC4 expression masks the trastuzumab binding epitopes of HER2, resulting in steric hindrance of the interaction between this antibody and its therapeutic target. Preclinical studies with JIMT-1 trastuzumab-resistant breast cancer cell line reported that the level of MUC4 protein is inversely related with the trastuzumab binding capacity (
10), and showed that knockdown of MUC4 is able to restore trastuzumab sensitivity of breast cancer cells in vitro. In the same direction, additional pre-clinical studies demonstrated that breast cancer cell lines are able to acquire resistance to trastuzumab through the upregulation of a cleaved form of the MUC1 protein, named MUC1*. Of note, such resistance can be reverted by MUC1* antagonists. A similar mechanism of resistance to trastuzumab has been reported for the protein CD44 ligand hyaluronan, that masks the HER2 extracellular domain. Of note, as in the case of MUC proteins, inhibition of CD44 by RNA interference can restore sensitivity to trastuzumab (11).

Truncated receptors
Cleavage of the full-length 185 kDa HER2 protein by matrix metalloproteases produces a 110 kDa extracellular domain (ECD), which is released into cell culture media or circulates in the serum in vivo, and a 95 kDa (p95 HER2) NH2-terminal-truncated membrane-associated fragment with increased kinase activity. The p95 HER2 protein may be generated by other mechanisms as alternative RNA-splicing or initiation of translation within the HER2 sequence. This p95 HER2 protein expression has been found in approximately 30% of HER2 positive breast cancers and cell lines. Of potential importance, trastuzumab blocked HER2 ECD proteolytic cleavage in vitro(12).

Elevated serum levels of HER2 ECD correlate with poor prognosis and trastuzumab resistance in patients with advanced breast cancer (13, 14). MCF-7 and T47D breast cancer cells transfected with p95HER2 show increased activation of Akt and MAPKs; treatment with trastuzumab for 2 hours resulted in no inhibition or a slight increase in the phosphorylation levels of HER2 in these cells. In 46 tumors from patients treated with trastuzumab alone or in combination, Scaltriti et al. found that p95 HER2 expression was strongly associated with trastuzumab resistance; only one of nine (11.1%) patients with tumors expressing p95 HER2 responded (with a partial response) to trastuzumab. HER2-targeted monoclonal antibodies could bind to circulating ECD, competing away the binding to membrane-bound HER2. The predictive role of elevated baseline ECD prior to treatment is not well defined. In one study, elevated HER2 ECD levels predicted favorably for response to trastuzumab and docetaxel, but other studies showed limited predictive value in this setting. Interestingly, declining levels of circulating HER2 ECD correlate with improved disease free survival in several studies. A meta-analysis of 8 clinical trials revealed that patients whose HER2 ECD levels declined by at least 20% in the first few weeks after initiation of trastuzumab-based therapy had improved disease-free and overall survival compared with patients whose HER2 ECD levels did not drop (15).

A recent study suggests that truncated forms of HER2 are the result of alternative initiation of translation from different methionines within the HER2 sequence, which are referred to as C-terminal fragments of HER2. The C-terminal fragment of HER2 may be present in the cytosol and behave as a transcriptional factor p95HER2-expressing breast cancer cells can activate potent growth and pro-survival signals through p95HER2–HER3 heterodimers (16).

Ligand excess and heterodimers
Growth factor ligands of EGFR, HER3, or HER4 reduced growth inhibitory effect of trastuzumab by 57, 84, and 90 percent, respectively. Trastuzumab causes destabilization of constitutive HER2-HER3 complexes, blocking downstream PI3K/Akt signaling. Heregulin (HN) secretion prevents the trastuzumab-dependent disruption of HER2/HER3 complexes, maintaining PI3K activity and conferring resistance. Overexpression of other EGFR or HER3 ligands, such as betacellulin, transforming growth factor (TGF)-a, EGF, heparin-binding EGF (HB) may confer resistance to trastuzumab via the constitution of heterodimers HER2/HER3 or HER1/HER2 (17, 18).

HER2 has no ligand and it is spontaneously in on-position. EGFR and HER2 coexist in 30% of breast cancers, conferring trastuzumab resistance via HER2-EGFR heterodimer, which signals also via EGFR. HER3 lacks a functional kinase domain; its dimerization with HER2 constitutes the most potent among all possible HER family heterodimers. The constitution of these heterodimers led to impair the degradation of trastuzumab-HER2 complex by the ubiquitin-proteasome. The combination of pertuzumab, which prevents HER2-HER3 heterodimerization, with trastuzumab produced responses and clinical benefit in 24% and 50% of patients progressing on trastuzumab (19), and studies are now evaluating the effect of combining with trastuzumab when resistance to pertuzumab occurs.

Interaction with other receptors
In the preclinical model of SKBR3 HER2-positive BC cells transfected with IGF-1R gene and exposed continuously to IGF-1, cell growth and proliferation continue even in presence of trastuzumab under the undisputed stimulation of downstream PI3K/Akt and MAPK pathways. Of note, the addition of IGF-binding protein 3 is able to restore the original sensitivity to trastuzumab (20). IGF-1R physically interacts with and phosphorylates HER2 receptor in trastuzumab-resistant cells, but not in trastuzumab-sensitive parental cells. Moreover, in trastuzumab-resistant cells, formation of HER2-IGF-1R heterodimers is able to phosphorylate and activate HER2, via stimulation of IGF-1R by IGF-1-like ligands (21). In the trastuzumab-resistant cell line, the combination of trastuzumab with an IGF-1R inhibitor is able to decrease the levels of p27kip1 and cell proliferation (22).

Met receptor is able to promote the breakdown of cell-cell junctions and to enhance cell invasion in tumors overexpressing HER2. Trastuzumab upregulates Met expression, which, in turn, by cross talk activation of HER3, favors the onset of trastuzumab resistance (23). Cell growth inhibition induced by trastuzumab was increased by c-Met inhibitor.

Downstream signaling
Trastuzumab resistance has been linked to activation of the phosphoinositol 3-kinase (PI3K) pathway. Constitutive activation of PI3K most frequently occurs via two mechanisms: loss of function PTEN (25%), or activating mutations in PIK3CA (25%) (24,25). Furthermore, trastuzumab-resistant cells derived from the BT474 HER2-overexpressing breast cancer line demonstrated elevated levels of phosphorylated Akt and Akt kinase activity compared with parental cells (26). PIK3CA mutations and PTEN loss are not mutually exclusive. The resistance disappears if the PI3-kinase activity is blocked: PI3K antagonists inhibit growth and survival of trastuzumab-resistant PI3K-mutant breast cancer cells (27). However, PI3K blockade may not be enough because, as compensation, HER2 may activate pERK. Activation of PI3K causes also mislocalization of the cyclin-dependent kinase inhibitor p27 in the cytoplasm, whereas loss of p27 expression confers resistance to trastuzumab (28).

It has been described an involvement of heat-shock proteins (HSP) 27 and 90 in HER2 pathway: a HSP-90 inhibitor showed to block proliferation in trastuzumab resistant cells (29); inhibition of HSP-27 via RNA interference restored the sensitivity to trastuzumab (30).

Alternative pathways
Treatment with trastuzumab triggers expression of cell fate regulatory protein and survival factor Notch-1 as compensatory feedback. Addition of a Notch inhibitor to drugs anti-HER2 causes a cooperative antiproliferative activity (31). Chemokine receptor CXCR4 and b1 integrin (CD29) may also play a role in the clinical resistance to HER2-targeted therapy while studies are ongoing to evaluate the role of the fibroblast growth factors and their receptors. In preclinical models (32, 33), as well as in clinical experience, resistance to trastuzumab is mediated by enhanced estrogen receptor (ER) signaling via up-regulation of the ER.

Lapatinib is a small molecule, a HER1 and HER2 dual tyrosine kinase inhibitor. Initial studies were performed based on preclinical data showing its potential activity in the subset of p95HER2-overexpressing, trastuzumab-resistant breast tumors. Lapatinib is registered in combination with capecitabine for metastatic BC progressing on antracycline, taxane and trastuzumab therapy (
34) and is approved in combination with letrozole in post-menopausal patients with HER2/ER-positive metastatic breast cancer not previously treated with aromatase inhibitors and or trastuzumab (35).

Mechanism of resistance to lapatinib. Response to lapatinib as single agent ranges from 24% to 1–4% in untreated and heavly pretreated patients respectively (36, 37). The main mechanisms of resistance to lapatinib include alterations in kinase domain of HER2; alternative signaling through unblocked heterodimers; activation of downstream signaling or compensatory pathways (Fig. 1, Tab. 2).

Table 2 Table 2
Mechanisms of resistance to lapatinib.

Mutations in kinase domain: similarly to what described in lung cancer for EGFR inhibitors, tyrosine kinase inhibitors (TKIs) targeting HER2 may induce secondary mutations in the receptor, leading to TKI resistance. In vitro studies identified and mapped point mutations to the HER2 kinase domain, clustering in the NH2-terminal kinase lobe and hinge region, and conferring resistance to the small-molecule TKI lapatinib by multiple mechanisms, including direct steric interference and restriction of conformational flexibility. HER2 T798I imparts the strongest lapatinib resistance effect.

Heterodimers: the constitution of HER2/HER3 heterodimer represents a mechanism of resistance common to all TKIs. Although they could decrease the kinase activity of HER2, TKIs are not able to block the recruitment of HER3 by PI3K and AKT pathway (38).

Downstream signaling: phosphoproteome analysis of breast cancer cells resistant to lapatinib revealed an increase in the mTOR effector p70S6K1, suggesting that activation of mTOR pathway is an escape to HER2 blockade and encouraging the combination of everolimus with HER2/EGFR inhibitors (39).

Compensatory pathways: it has been shown that the TKR BRK is coexpressed and coamplified with HER2 in some breast cancer cell lines, suggesting that both receptors cooperate. Lapatinib has no action when this cooperation prevails (40). Finally, another mechanism of resistance to lapatinib is linked to the presence of crosstalk between HER2 and ER pathway (32).

Pertuzumab is a monoclonal antibody that binds to an epitope on the domain II of the HER2 ECD, distinct from the binding site of trastuzumab, that is on the domain IV. Preclinical and clinical data support the central role of the HER2-HER3 interaction in PI3K/Akt-mediated tumorigenesis in HER2-overexpressing breast cancers supporting the scientific rationale of developing this agent. Pertuzumab monotherapy has limited activity in heavily pretreated patients with HER2-positive MBC and in patients with HER2-negative MBC. Of note, in a phase II study of 66 patients with HER2-positive MBC who had progressed on prior trastuzumab, combination antibody treatment with pertuzumab and trastuzumab (without chemotherapy) led to an ORR of 24.2% (19).

Mechanism of resistance to pertuzumab. The main mechanisms of resistance to pertuzumab are: activation of downstream signaling or alternative and compensatory pathways (Fig. 1, Tab. 3).

Table 3 Table 3
Mechanisms of resistance to pertuzumab.

Heterodimers: pertuzumab has shown to be an effective inhibitor of HER1/HER2-driven signalling and growth, blocking Akt and ERK1/2 activity. However, pertuzumab also promotes a rapid formation of HER3/HER1 heterodimers providing a mechanism whereby EGF and HRGß1 stimulation could circumvent its growth inhibitory effect. Notably, the HER3/HER1 heterodimer levels are increased in pertuzumab-resistant cells, further suggesting that this heterodimer may play a central role in the rapid acquisition of resistance to this agent.

Downstream signaling: in a preclinical model, Akt activation correlates with insensitivity to pertuzumab, whereas inhibition of Akt correlates with response to pertuzumab. Compensatory pathways: pertuzumab doesn’t inhibit growth of tamoxifen-treated MCF-7TAMLT tumors. In contrast, pertuzumab inhibits estradiol plus fulvestrant-stimulated growth of MCF-7TAMLT tumors by 37.2%. These results suggest that tamoxifen stimulated growth of MCF-7TAMLT tumors in vivo might be independent of EGFR, HER2/neu, or HER3 pathways.

PI3K/Akt/mTOR inhibitors
As mentioned above, trastuzumab resistance can be linked to PI3K pathway activation via PTEN loss or PI3K mutation with constitutive activation of Akt/mTOR. This pathway has been studied as a therapeutic target in cellular and xenograft models. Resistance to trastuzumab and lapatinib can be reverted by PI3K/mTOR inhibitor (
41, 42). The addition of the mTOR inhibitor everolimus to trastuzumab in MBC paients heavily pretreated and progressed to trastuzumab has shown a 15% of response rate and a progression free survival of 4.1 months. Similarly, adding everolimus to trastuzumab and paclitaxel can restore sensitivity to trastuzumab and taxane, with a 25% of response rate (43).

Similar results have been reported for lapatinib, as pre-clinical models of lapatinib resistance, pharmacologic inhibition of mTOR with rapamycin or ridaforolimus increased lapatinib sensitivity and reduced phospho-Akt levels. Combining mTOR inhibitors with lapatinib resulted in supradditive inhibition of proliferation, reduced anchorage-independent growth, and reduced in vivo tumor growth of HER2-overexpressing breast cancer cells that have primary trastuzumab resistance (44).

Clinical trials have shown the efficacy of adding anti-HER2 therapies both in the metastatic and in the neoadjuvant setting: in MBC, adding another anti-HER2 drug has shown to be better than replacing the first one. The addition of lapatinib to trastuzumab in trastuzumab-progressed patients, shows a median progression-free survival of 12 weeks compared with 8.4 weeks of lapatinib alone (HR 0.77; 95% CI 0.6–1.0, p=0.029) (45). In the NeoALLTO trial, the pCR rate was significantly higher in the group given lapatinib and trastuzumab (51.3%; 95% CI 43.1–59.5) than in the group given trastuzumab alone (29.5%; 22.4–37.5) with a 21.1% difference (9.1–34.2, p=0.0001) (7). Same results have been shown in pertuzumab trials: in the CLEOPATRA, the median progression-free survival was 12.4 months in the trastuzumab group, as compared with 18.5 months in the pertuzumab plus trastuzumab group (hazard ratio for progression or death, 0.62; 95% CI 0.51–0.75; p<0.001) with a strong trend in overall survival in favor of pertuzumab plus trastuzumab (46). In the Neo-sphere trial, patients given pertuzumab and trastuzumab plus docetaxel had a significantly improved pathological complete response rate (45.8%, 95% CI 36.1–55.7) compared with those given trastuzumab plus docetaxel (29.0%, CI 20.6–38.5; p=0.0141), those given pertuzumab plus docetaxel (24.0%, CI 15.8–33.7) and women given pertuzumab and trastuzumab (16.8%, CI 10.3–25.3) (47).

Treatment with anti-HER2 agents works in breast cancer addicted to HER2 signaling until compensatory mechanisms arise overcoming this selective pressure. Resistance occurs when cancer cells switch their dependence from HER2 to another oncogenic pathway for proliferation and survival. Recently, it was suggested that addictive pathway not only generates oncogenic signals but triggers also latent tumor-suppressor signals. So, tumor-suppressive signals may be inhibited by targeted therapies, ie trastuzumab can block tumor-suppressive proteins at the same time in which it blocks the oncogenic protein HER2. Therefore, dormant tumor cells could be awakened by prolonged targeted drugs, as in the case of maintenance of anti-HER2 therapy (50). According to this hypothesis, tumor response to targeted therapy is the result of a dynamic process whereby a balance exists between proliferative and anti-proliferative signals in replicating and dormant cells. When balance fails, the “oncogenic shock” kicks in (51). On the basis of the above, the “drug sedimentation theory”, suggests to get advantage of using sequential and combination treatments with targeted agents to prevent resistance. Further to being just a comprehensive rational model based on preclinical data, this theory has also been recently demonstrated in the neoadjuvant setting with the combinations of trastuzumab and lapatinib/pertuzumab. Additional studies are however needed to select the most effective strategy to treat HER2 breast cancer phenotypes.

Thanks to Valentina Serpico for her contribution in picture realization.