- © 2006 by American Society of Clinical Oncology
Can Trastuzumab Be Effective Against Tumors With Low HER2/Neu (ErbB2) Receptors?
HER2/neu (erbB2) is a member of the erbB family of transmembrane receptor tyrosine kinases, which also includes the epidermal growth factor receptor (EGFR; erbB1), HER3 (erbB3), and HER4 (erbB4). Binding of ligands to the extracellular domain of EGFR, HER3, and HER4 induces the formation of kinase active homo- and heterodimers to which HER2 is recruited as a preferred partner.1 Although HER2 does not bind any of the erbB ligands directly, its catalytic activity can potently amplify signaling by erbB-containing heterodimers via increasing ligand binding affinity and/or receptor recycling and stability.2-5 Activation of the erbB network leads to receptor autophosphorylation in C-terminal tyrosines and the recruitment to these sites of cytoplasmic signal transducers that regulate cellular processes such as proliferation, differentiation, motility, adhesion, protection from apoptosis, and transformation.1 HER2 gene amplification has been reported in approximately 20% of metastatic breast cancers, where it is associated with poor patient outcome.6 Trastuzumab, a humanized monoclonal immunoglobulin G1 that binds the extracellular domain of HER2, has been shown to induce clinical responses in HER2-overexpressing breast cancers and prolong patient survival when combined with chemotherapy.7-10 These studies enrolled either predominantly or exclusively patients with breast cancer that overexpressed HER2 as measured by intense membrane staining with HER2 antibodies in the majority of tumor cells (3+ by immunohistochemistry [IHC]) or excess copies of the HER2 gene determined by fluorescent in situ hybridization.
Studies with human breast cancer cell lines and primary mammary tumors have shown constitutive phosphorylation of HER2.11,12 Thor et al13 examined the incidence of P-HER2 positivity in 816 primary breast cancers. Only 12% of HER2-positive cancers stained positively with an antibody that recognizes HER2 phosphorylated in Y1248; P-HER2 was not detected in any of the HER2-negative tumors.13 The molecular basis for this constitutive phosphorylation is not clear but is consistent with the reported ability of wild-type Neu, the mouse/rat homolog of HER2, to multimerize and become activated when present in cells at high density.14 Thus, it is generally accepted that spontaneous dimerization and activation of HER2 occurs in cancers with HER2 gene amplification. Another potential mechanism of HER2 phosphorylation is transactivation by ligand-bound EGFR, HER3, or HER4. Indeed, coexpression of the EGFR ligand transforming growth factor alpha (TGFα) and Neu in the mammary gland of transgenic mice markedly accelerates tumor onset and progression compared with mice expressing Neu or TGFα transgenes alone. In this study, TGFα × Neu bi-transgenic mice exhibited increased tyrosine phosphorylation of both EGFR and Neu.15 In an analysis of 807 patients with invasive breast cancers, 306 HER2-positive tumors also expressed EGFR by IHC. Ninety-seven percent of cancers with phosphorylated HER2 at Y1248 exhibited detectable EGFR and the combination of Y1248 P-HER2 together with co-overexpression of HER2 and EGFR was associated with the shortest patient survival.16 Finally, HER2 is overexpressed in a cohort of non–small-cell lung cancers and increased HER2 gene copy number has been associated with therapeutic response to EGFR tyrosine kinase inhibitors.17,18 These reports provide evidence of EGFR-HER2 cross-talk and of the ability of HER2 to amplify ligand-activated EGFR signals.
Immunohistochemical studies have shown HER3 expression and overexpression in over 50% of ductal carcinoma in situ and invasive breast cancers.19-23 Although kinase defective, HER3 can be phosphorylated by EGFR or HER2.1,24 Phosphorylated HER3 can couple to the phosphatidylinositol-3 kinase (PI3K) pathway directly, whereas EGFR and HER2 cannot.25 HER2/HER3 heterodimers are potently transforming24,26 and both HER2 and HER3 are frequently associated or coexpressed in human breast cancer cell lines, mouse transgenic tumors, and primary breast cancers.11,19,20,27,28 In HER2-overexpressing tumor cells, both trastuzumab and EGFR kinase inhibitors inhibit the basal phosphorylation of HER3 and its association with HER2 and with PI3K.12,29,30 Finally, high expression of HER3 predicts early escape from trastuzumab therapy.31 These data imply that HER3 plays a key role in breast tumor progression mediated by the erbB network, and inhibition of HER3 phosphorylation and/or its cross-talk with HER2 is required for the antitumor effect of HER2 signaling inhibitors. Whether incorporation of markers such as HER3, P-HER2, EGFR, EGFR/HER2 dimers, HER2/HER3 dimers or others that may inform the level of erbB pathway engagement, will allow the identification of patients that respond or escape trastuzumab therapy remains to be investigated.
Structural studies using recombinant erbB receptor ectodomains, as well as experiments with cell lines and HER2 monoclonal antibodies that recognize different receptor epitopes, indicate that trastuzumab binds in a region not involved in receptor dimerization and thus is unable to block ligand-induced EGFR/HER2 and HER2/HER3 heterodimers.32-34 These data support the speculation that high levels of HER2-containing heterodimers may be a potential marker of resistance to trastuzumab. In support of this, the EGFR ligand TGFα has been shown to abrograte trastuzumab action against antibody-sensitive breast cancer cells.29,35 Furthermore, high expression of EGFR is associated with early escape from trastuzumab therapy.31
These cumulative data lead to three generally accepted premises in this field. First, HER2 is only a pathogenic oncogene in those tumors with HER2 gene amplification and/or high HER2 protein overexpression as defined by (3+) IHC. Second, trastuzumab either alone or in combination with chemotherapy is only effective against those same tumors but not against cancers with low (physiological) levels of the gene and/or tumors without detectable HER2 protein by IHC. Third, trastuzumab does not impair the recruitment of HER2 for dimerization with ligand-activated HER3 or EGFR and is, therefore, not a complete inhibitor of the erbB network in cancers with amplified HER2 signaling. Some of these premises are challenged by the work presented by Menendez et al36 in this issue of the Journal of Clinical Oncology.
In this article, MCF-7 human breast cancer cells stably overexpressing the HER3 ligand heregulin and single-copy low levels of HER2 were resistant to the anticancer drugs cisplatin, paclitaxel, vincristine, and fluorouracil. Cotreatment with trastuzumab completely reversed drug resistance. Treatment with trastuzumab inhibited HER2 and HER3 phosphorylation and blocked signaling by PI3K/Akt and Erk in heregulin-overexpressing cells, but not in control cells. Despite its antisignaling effects, the antibody by itself did not inhibit cancer cell growth. The experiments addressing the antibody effects on signaling were done in the absence of chemotherapeutics. Interestingly, the cytotoxic effect of cisplatin was enhanced by cotreatment with trastuzumab in human breast cancer cells that naturally overexpress heregulin and also contain single copy HER2, so called HER2 negatives. This sensitizing effect was mimicked by transfection of an antisense cDNA against heregulin. Unfortunately, confirmatory experiments in vivo were not included. The authors conclude that blockade of HER2-dependent, heregulin-induced autocrine action with trastuzumab triggers receptor-enhanced chemosensitivity in the absence of HER2 overexpression, and, therefore, breast cancers that overexpress heregulin and P-HER2 regardless of total HER2 levels would benefit from therapy with trastuzumab and chemotherapy.
Induced overexpression of heregulin results in enhanced transformation and tumorigenicity37-40 and, in the article discussed herein, led to relative resistance to anticancer drugs. Treatment with trastuzumab reversed therapeutic resistance in vitro. These data suggest that a transforming heregulin-activated HER2-HER3 complex with low levels of HER2 is indeed HER2-dependent and is thus a rational therapeutic target of HER2-specific inhibitors. Further circumstantial evidence in support of this notion is presented by Menendez et al36 in an analysis of 189 invasive breast cancers: only three (5%) of 57 specimens with detectable HRG levels were HER2 overexpressors but 38 (67%) of 57 specimens showed staining with the Y1248 P-HER2 antibody, implying that an excess of the HER3-activating ligand may dispense with the need of HER2 overexpression by inducing high levels of HER2 dependent signal amplification. This concept of ligand-activated nonamplified receptor tyrosine kinases driving cancer progression is not unprecedented. For example, a type of sarcoma called dermatofibroma protuberans, in which a t(17,22) chromosomal translocation leads to constitutive platelet-derived growth factor ligand production is exquisitely sensitive to the platelet-derived growth factor receptor kinase inhibitor imatinib.41 Prostate cancers that become androgen-independent acquire colocalization of TGFα and EGFR.42 High levels of TGFα are associated with poor survival in squamous cancers of the head and neck independent or EGFR overexpression.43 Interestingly, cetuximab, a monoclonal antibody that blocks EGFR ligand binding, has activity as a single agent against cancers of the head and neck.44 Therefore, considering the potent kinase activity of HER2, it is not inconceivable that cancers with overall low HER2 but a high proportion of (trans)activated receptors, are HER2 dependent.
The results of Menendez et al36 disagree in part with those reported by Agus et al,32 in which trastuzumab was unable to block signaling induced by exogenous heregulin in MCF-7 cells. This inconsistency suggests differences on the mechanisms by which soluble heregulin and membrane-anchored heregulin (in MCF-7, MDA-231, and HS578T cells in the article being discussed) interact with erbB receptors, as already suggested by Yuste et al.37 Although not shown, these results also imply that trastuzumab can disrupt HER2/HER3 heterodimers induced by the autocrine action of membrane-anchored heregulin.
Menendez et al, propose that breast cancers that overexpress heregulin and P-HER2 but contain low overall HER2 receptors would benefit from therapy with trastuzumab and chemotherapy. Neither of these biomarkers alone or in combination has been tested previously in any of the large randomized studies testing the efficacy of trastuzumab. In a large study examining the prevalence of Y1248 P-HER2 positivity, only 12% of HER2-positive cancers were P-HER positive while phosphorylated HER2 was not detected in any of the HER2-negative tumors.13 In this study, any membrane staining with the CB11 erbB2 antibody was considered positive, thus implying that in HER2-negative tumors, Y1248 P-HER2 is not detectable. This leaves tumors with low and intermediate HER2 protein levels and no HER2 gene amplification, those scored as 1+ or 2+ by IHC, as cohorts where heregulin and/or P-HER2 could be detectable or overexpressed. Unfortunately, there are no data in this report about the frequency of heregulin and/or P-HER2 detection in HER2-positive breast cancers that are 1+ or 2+ by IHC. Therefore, the prevalence of the molecular signature proposed by Menendez et al36 for treatment with trastuzumab requires further definition in a larger number of cancers to support the feasibility of their proposal. In addition to measuring heregulin and HER3, this analysis should also incorporate detection of other autophosphorylation sites in HER2. Inclusion of additional site-specific P-HER2 antibodies may well help explain the rather low 12% positivity of Y1248 P-HER2 among HER2-positive tumors in this report by Menendez and the larger study by Thor et al.13 If heregulin and P-HER2 positivity is limited to HER2 IHC 2+ cancers, the proposition of treating these tumors with trastuzumab would weaken significantly as these tumors have already been shown to derive little to no benefit from therapy with trastuzumab either alone or in combination with chemotherapy.7,8
In summary, the article by Menendez et al36 challenges the current dogmas in this area of molecular therapeutic investigation. It reaffirms the role of ligand-induced signal amplification in mediating receptor tyrosine kinase dependence and, thus, the testable efficacy of HER2 inhibitors against tumors with single-copy, low (but activated) HER2 levels. An important practical implication from this report is the need of incorporate erbB ligands and other biomarkers indicative of receptor phosphorylation/activation to the molecular profile obtained in tumors/patients enrolled in trials with drugs targeted against EGFR, HER2, or HER3. A more complete molecular and biochemical characterization of the erbB ligand/receptor network in primary cancers should ultimately help us to predict response or lack of response to partial inhibitors of this oncogenic network, such as trastuzumab.
Author's Disclosures of Potential Conflicts of Interest
The author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
| Authors | Employment | Leadership | Consultant | Stock | Honoraria | Research Funds | Testimony | Other |
|---|---|---|---|---|---|---|---|---|
| Carlos L. Arteaga | Genentech (C) |
Dollar Amount Codes (A) < $10,000 (B) $10,000-99,999 (C) ≥ $100,000 (N/R) Not Required
Acknowledgments
Supported by National Cancer Institutes R01 Grants No. CA62212 and CA80195, Breast Cancer Specialized Program of Research Excellence (SPORE) Grant No. P50 CA98131, and Vanderbilt-Ingram Comprehensive Cancer Center Support Grant No. P30 CA68485.
Footnotes
-
published online ahead of print at www.jco.org on July 17, 2006.
