Is the Type I Insulin-Like Growth Factor Receptor a Therapeutic Target in Endometrial Cancer?

How do we discover therapeutic targets in cancer? Because we know that cancer is a disease of accumulated genetic abnormalities, an ideal cancer target should be an altered protein with aberrant function resulting from a genomic abnormality. Certainly, understanding the actions of oncogenes and tumor suppressors has guided us towards new drugs, with trastuzumab for breast cancer and imatinib for chronic myelogenous leukemia being the most obvious.

However, proteins that are not genetically altered and retain their normal function can also be excellent targets. Estrogen and androgen receptors are examples of proteins that are almost never altered by genetic mechanisms, and their continued function is critical for the development and treatment of breast and prostate cancer, respectively. Moreover, these proteins must be activated by ligand to function; both inhibition of ligand production and disruption of receptor function can be effective therapies.

Are there other similar targets for cancer therapy? In this issue of Clinical Cancer Research, McCampbell et al. report that the type I insulin-like growth factor receptor (IGF-IR) is highly expressed and activated as judged by its phosphorylation status in hyperplastic endometrium and type I endometrial cancer (1). Is this enough to suggest IGF-IR is a therapeutic target in endometrial cancer?

Although early-stage endometrial cancer is highly curable by surgery alone, locally invasive and metastatic disease has a poorer prognosis. Unlike other types of cancers, such as breast cancer and colon cancer, where targeted therapies against specific protein receptors have been applied clinically, no targeted therapy has been developed against endometrial cancer. At least part of the reason could be due to the lack of knowledge of the molecules and proteins that control the growth and development of endometrial cancer. To date, for type I endometrial cancer (estrogen dependent), the identified genetic abnormalities include microsatellite instability, loss of PTEN expression, and gene mutations of PTEN, K-ras, and β-catenin. Whereas for type II, non–estrogen-dependent endometrial cancers, distinct genes are involved, including p53 mutations and human epithelial growth factor receptor-2 (HER2) overexpression (summarized in review ref. 2). All of these events involve genetic alterations that cause abnormal protein function. McCampbell et al. forward the idea that “normal” IGF-IR and its signaling components may play a role in type I endometrial cancer.

IGF-IR is a transmembrane receptor tyrosine kinase composed of two α subunits and two β subunits (3). Unlike HER2, where overexpression can trigger its activation, the activation of IGF-IR requires binding to either of its ligands, IGF-I or IGF-II. Upon ligand binding to the extracellular α subunits, the kinase domains within the two β subunits are activated, and certain tyrosine residues are phosphorylated, which leads to the subsequent recruitment and activation of adaptor molecules, such as insulin receptor substrates 1 and 2. In addition to IGF-IR, hybrid receptors composed of one αβ subunit from insulin receptor and one αβ subunit from IGF-IR also exist (4). As a result multiple downstream signaling pathways are activated, including the mitogen-activated protein kinase pathway and the phosphatidylinositol 3-kinase pathway, which lead to cell proliferation, motility, and metastasis (5). The IGF-IR activity can be attenuated by receptor internalization and/or phosphatase activity. After activation, the mitogen-activated protein kinase pathway and phosphatidylinositol 3-kinase pathway can also be attenuated at multiple levels. For example, the lipid phosphatase PTEN can dephosphorylate PIP-3, which is the second messenger produced by phosphatidylinositol 3-kinase, and therefore can deactivate the phosphatidylinositol 3-kinase pathway (Fig. 1).

What does the IGF system do in normal growth and development? Mice with a gene deletion of either IGF-I or IGF-IR have severe growth abnormalities (6). Humans with functional disruptions in the IGF-I or IGF-IR have similar growth abnormalities (7, 8), supporting the critical role for this system in the growth of essentially every normal organ. If this receptor system plays a role in normal growth, then it could have a role in malignant growth as well. It has been well documented that IGF-IR and its signaling components are critical regulators of certain types of cancer, including breast, colon, prostate, liver, myeloma, and melanoma malignancies (9). Several anti-IGF-IR agents, including antibodies and small molecules inhibitors, have been developed, and some of them are in clinical trials against myeloma, breast and prostate cancer (10).

IGF-IR system has been implicated in the normal development of endometrium. IGF-I and IGF-II are expressed in the endometrial stromal cells, and their expression is associated with endometrial differentiation (11). IGF-IR is expressed mainly in endometrial epithelial cells, and in a lesser extent, in stromal cells (12). Mouse stromal IGF-I gene expression can be stimulated by estrogen (13), and IGF-I increases the bromodeoxyuridine uptake in endometrial stromal cells (14).

The reports on the connection between IGF-IR and endometrial cancer have been sparse. IGF-IR mRNA and protein expression has been detected in endometrial tumor samples (15–17). IGF-I and II enhance the proliferation of a human endometrial carcinoma cell line, ECC-1 (18). McCampbell et al. are the first to show a correlation between high expression and activation of IGF-IR in complex atypical hyperplasia (CAH) and endometrial cancer.

In a total of 44 CAH and endometrial cancer samples, the IGF-IR transcript levels were significantly higher than in the 17 proliferative or secretory-phase endometrium samples. Accordingly, they observed higher expression of IGF-IR protein in both CAH and endometrial cancer. Interestingly, the levels of other signaling components, such as IGF-I, IGF-II, and IGF-binding proteins, were not elevated compared with normal tissues. It is unclear in CAH and endometrial cancer what protein up-regulates IGF-IR gene expression. In general, IGF-IR genes can be regulated by many growth factors and hormones (3). For example, activation of ERα up-regulates the expression of IGF-IR (19). In addition, the WT1 Wilms' tumor suppressor gene has been shown to repress IGF-IR gene expression both in vitro and in vivo (20). Decreased WT1 levels are correlated with up-regulation of IGF-IR gene expression in Wilms' tumor, benign prostatic hyperplasia, and breast cancer (20–22). It would be interesting to determine if WT1 levels decrease in CAH and endometrial cancer in future studies.

McCampbell et al. also compared the phosphorylation status of IGF-IR between proliferative-phase endometria and endometrial CAH and endometrial cancer. IGF-IR was highly phosphorylated in endometrial CAH and endometrial cancer but not in proliferative endometrium. As mentioned earlier, high IGF-IR expression alone does not trigger its activation. IGF-IR need to be activated by its ligands IGF-I and IGF-II (9). As shown, IGF-I and IGF-II are abundant in the tissue and could be the source of receptor activation. In addition, the constitutive phosphorylation/activation could be due to the lowered levels of phosphatase activity with an inability to shut off IGF-IR signaling or due to an abnormality in intracellular trafficking of the receptor. It is noteworthy that the antibody used by these investigators can also recognize phospho-insulin receptor, and it is possible that insulin receptor or IGF-IR/insulin receptor hybrids could contribute to the high level of activation detected. Therefore, to exclude the status of phosphorylated insulin receptor in tumor samples, a more thorough study would involve a characterization of both insulin receptor and IGF-IR. Immunoprecipitation could be used to distinguish between the activation states of the two receptors. This is especially important because of the known link between obesity, insulin resistance with higher circulating insulin levels, and endometrial cancer. In addition, IGF-II may also signal through the insulin receptor (23).

McCampbell et al. observed no correlation between the loss of PTEN expression and increased activation of IGF-IR in endometrial hyperplasia, which suggests that the regulation of PTEN expression and activation of IGF-IR are through distinct signaling pathways. However, as a consequence of loss of PTEN expression, the Akt signaling activated by IGF-IR was further enhanced. Accordingly, McCampbell et al. observed that the simultaneous loss of PTEN expression and increased IGF-IR activation in hyperplasia was associated with an increased incidence of endometrial cancer. These data suggest that in PTEN-deficient cells, upstream inputs to phosphatidylinositol 3-kinase and Akt may still be important with IGF-IR potentially driving this pathway. It is tempting to speculate that activation of IGF-IR signaling could be an early event in the progression of normal cells to hyperplasia with the loss of PTEN pushing cells towards frank malignancy.

The paper by McCampbell et al. provides data to support a role for IGF-IR signaling as a therapeutic target in endometrial cancer. However, additional data need to be generated to prove that IGF-IR function is relevant in this disease. First, the sample size reported here is relatively small. Conformation of the relevance of this pathway could be obtained by studying additional samples. Second, more biological studies are needed to investigate whether IGF-IR signaling causes malignant transformation of the endometrium. If it does, this would support a role for IGF targeted therapies in the prevention of the disease. Certainly, animal models designed to overexpress IGF-IR in the endometrium could begin to address this possibility.

Fortunately, drugs designed to disrupt IGF-IR signaling are rapidly being developed. Clinical correlation between gene/protein expression and prognosis can point us in the direction of new targets in cancer. Certainly, the correlative studies between HER2 expression and breast cancer outcome were the first clues that disruption of HER2 function might have therapeutic value in breast cancer (24, 25). Given that IGF-IR activation correlates with the development of endometrial cancer, testing anti-IGF-IR therapies in this disease is rational. Direct evidence that disrupting IGF-IR signaling affects clinical outcome is the clearest way to show the value of the target.