Molecular oncology. Causes of cancer and targets for treatment. Eds Edward P. Gelmann et al., Cambridge University Press (2014)
Insulin and insulin-like growth factors (IGFs) are potent mitogens, and the hypothesis that their receptors are important therapeutic targets in oncology has received considerable attention (reviewed in 1–5). In the last decade, more than 20 drug candidates that target IGF receptors or both insulin and IGF receptors have been developed. Of these, at least 12 have been taken forward to clinical trials.
The rationale for drug development in this area included clinical and epidemiologic evidence (for example 6,7) that circulating levels of insulin and/or IGFs are related to cancer risk and/or cancer prognosis, as well as laboratory studies (for example 8) which demonstrated that interfering with signaling had inhibitory effects on neoplastic behavior. Seminal studies (9) from the laboratory of Renato Baserga showing a requirement for presence of the IGF-I receptor for transforming activity of a variety of oncogenes also contributed to the rationale. This research was followed by laboratory studies of drug candidates that demonstrated activity (for examples 1,2,4,5), which then led to clinical trials. In retrospect, however, it must be recognized that most pre-clinical studies of drug candidates showed benefit in experimental cancer models engineered to be IGF-IR driven, or in models chosen specifically because they were sensitive to the drugs, with relatively little attention being given to studies of tumor or host characteristics that predicted activity.
Evidence from population studies are of particular interest; it is rare for circulating ligand levels for a tyrosine kinase receptor to be related to cancer risk or prognosis, but this has been shown for insulin and IGF-I (for example 6, reviewed in 1,4,5). For some tissue growth factors, such as EGF or PDGF, circulating levels may have no meaning; although the peptides are detectible, their levels are not regulated and have no physiologic role, as they simply “leak” into the circulation from the tissues where they are expressed. Insulin and IGFs, in contrast, have credentials both as tissue growth factors and as circulating hormones, and their circulating levels are regulated by complex physiologic control systems, and vary from person to person according to a variety of genetic and lifestyle factors.
Early findings (for example 6,10) that indicated that individuals with higher circulating levels of IGF-I are at increased risk for common cancers have in some cases been confirmed in subsequent studies (11,12), but for some disease sites, data are inconsistent (13–15). A poorly understood trend noted in certain data sets is that high levels of insulin (or c-peptide, often used as a surrogate) tend to be more useful in predicting prognosis among cancer patients than cancer risk in populations, while variation in IGF-I levels between individuals is more useful in predicting cancer risk than in determining prognosis after diagnosis. A report (16) of unexpectedly low cancer incidence in an “outlier” population with a mutation that drastically lowers IGF-I levels is consistent with the prior reports showing cancer risk varying within the broad “normal” range of IGF-I.
Recent studies of cancer risk and/or cancer mortality among diabetics raise many questions, but are nevertheless relevant to targeting the insulin/IGF-I receptor family in neoplasia. There is strong evidence that certain cancers are more common among type II diabetics (reviewed in 17), and the basis for this is a matter of ongoing research. There also is direct evidence for an association between hyperinsulinemia and adverse prognosis in breast cancer (18). Some experimental studies support the view that the hyperinsulinemia of type II diabetes is a key mediator of this effect, as suggested by increased activation of insulin receptors (and also increased glucose uptake; 19) by neoplasms in models of diet-induced hyperinsulinemia (20,21). If this is the case, then hyperinsulinemic cancer patients with insulin-responsive tumors might be particularly appropriate for therapies that reduce insulin levels and/or reduce signal transduction through the insulin receptor. However, other aspects of the pathophysiology of type II diabetes, such as elevations in inflammatory cytokines (22) or even hyperglycemia itself may also be important mediators of the adverse effect of type II diabetes on cancer risk and/or prognosis, so caution is required in attributing the diabetes–cancer association exclusively to insulin. While hyperglycemia is a candidate risk factor, experimental systems showing that type I diabetes, with hypoinsulinemia and hyperglycemia, is associated with reduced rather than increased tumor growth, arguing against this possibility (23). Thus, important epidemiologic research associating hyperglycemia with cancer risk (24,25) does not necessarily imply a causal relationship; hyperglycemia in these studies may serve as a surrogate for other pathophysiologic mediators. Research concerning diabetes–cancer interactions has also provided evidence that metformin treatment of diabetics is associated with decreased cancer risk (reviewed in 17,26,27). The mechanisms involved here may include the reduction of hyperinsulinemia by this agent, but pre-clinical evidence suggests that is only one of several relevant actions (reviewed in 17,26,27).
If insulin can stimulate aggressive behavior of certain cancers, is it possible that insulin therapy for type II diabetes might have an adverse impact on cancer behavior? Theoretically, this would be predicted, particularly as insulin levels associated with conventional subcutaneous insulin administration are higher than those of non-diabetics (28,29). This arises not only because insulin is given at doses adequate to improve blood glucose under conditions of insulin resistance, but also because the amount of insulin required to impact the liver via the subcutaneous route is considerably higher than the amount required via physiologic secretion at pancreatic cells. Many studies have attempted to address this issue (reviewed in 27), but data to associate insulin therapy in general (or any particular insulin analog) with cancer risk are not conclusive. These studies are obviously difficult to carry out due to many potentially confounding variables, and to issues of dose, duration, and cotreatment. Some studies raise concern and certainly justify furher research (30).
Progress in physiology and pathophysiology
The fundamental aspects of relevant physiology and pathophysiology have been reviewed previously (1,4,5), but certain aspects deserve emphasis in view of recent recognition of potential relevance to therapeutic targeting.
The insulin/IGF receptor family
It was initially surprising to some endocrinologists that insulin and IGF-I receptors, previously well characterized on the normal tissues that classically are known to be responsive to their respective ligands, are also commonly expressed at significant levels on cancer cells derived from almost all organs (for example 31,32). However, this is now accepted, and it is recognized that from an evolutionary perspective insulin-like signaling in control of cellular proliferation is older than the relatively recent specialized functions in regulation of carbohydrate metabolism (insulin) or skeletal growth (IGF-I).
While investigators sometimes find it helpful to compartmentalize their work and consider insulin and IGF receptors as distinct topics in oncology, it is probably more accurate to consider this receptor family in a unified fashion. Review of cancer gene-expression databases reveals that most cancers express both the gene encoding the insulin receptor and the gene encoding the IGF-I receptor; thus most individual cancer cells (or epithelial cells at risk for transformation) likely express both receptors, and studies of one receptor alone may be incompletely reflect physiology. Even more fundamentally, it is now understood that when both receptors are expressed, it is the rule rather than exception for “hybrid” receptors to be present on the cell surface (for example 33, reviewed in 34). Recalling that both insulin and IGF-I receptors are heterodimers, where each “half receptor” is composed of an chain and a chain, the assembly of “half receptors” into “holoreceptors” appears to occur on the cell surface in a manner that does not necessarily favor pairing of identical “half receptors,” so that the formation of so-called “hybrid” receptors is common. In fact, if both gene products are translated at similar rates, hybrid receptors may be the dominant species on the cell surface. While certain normal tissues express one of the receptor genes much more abundantly than the other (e.g. the liver is a classic insulin-responsive tissue, with abundant insulin receptors, but few IGF-I receptors), this is not generally true for neoplastic cells. In general, therapeutic antibodies directed against the IGF-I receptor also block hybrid receptors, but detailed studies have not been performed with all drug candidates.
Additional complexity concerns the two isoforms of the insulin receptor (34), which arise due to alternate splicing. It is now recognized that one isoform (“IR-A”) is preferentially expressed by cancers, and although it is structurally similar to the “IR-B” isoform, it has potentially important differences in ligand binding. While the IR-B isoform is insulin-specific, IRA also binds IGF-II, which may be significant for those neoplasms where pathophysiology involves an autocrine IGF-II loop.
Considering jointly the complexities of hybrid receptors and insulin receptor isoforms, the insulin-IGF-I receptor family actually has six potential members, including each combination by which “half receptors” of IGF-IR, insulin-receptor isoform A, and insulin receptor isoform B can combine with a partner to generate the receptor complex. As we have seen, it is common for several of these receptor species to be expressed on a single cell. Apart from this, a receptor known as the IGF-II receptor must be considered. This receptor binds only IGF-II and is structurally distinct from the other family members, with the most obvious difference being the lack of an intra-cellular tyrosine kinase domain. Most investigators believe this receptor does not transduce a mitogenic signal, and actually acts as a growth inhibitor or tumor suppressor by binding IGF-II and sequestering it away from the IR-B or IGF-I receptors (or their hybrids), thus reducing IGF-II bioactivity (35).
A recent paper (Evdokimova V et al. IGFBP7 Binds to the IGF-1 Receptor and Blocks Its Activation by Insulin-Like Growth Factors. Sci Signal. 2012 Dec 18;5(255) provides evidence that one of the weak IGF ligand-binding proteins, known as IGFBP-7, or mac25, actually functions as an endogenous blocker of IGF-I receptor activation, adding further complexity to the regulation of IGF signaling.
Variation in levels of receptor expression between transformed and normal tissue remains a topic of active investigation, although it is clear that gene amplification with greatly increased receptor number per cell leading to ligand-independent activation, as seen with HER2-neu, is rarely seen. Variation in receptor levels between tumors has been noted, and there may be relationships between receptor level and disease prognosis or response to drugs that target this receptor family (36,37). It is important to recognize that standard immunostaining approaches with conventional anti-receptor antibodies do not discriminate between all family members. Most insulin receptor antibodies detect both the IR-A and IR-B isoforms. Furthermore, immunoreactivity with an anti-insulinreceptor antibody may signify presence of insulin receptors and/or hybrid receptors, and similarly immunoreactivity with an anti-IGF-I receptor antibody may signify presence of IGF-I receptor and/or hybrid receptors.
The first steps in the signaling networks downstream of the insulin receptor and IGF-I receptor are similar, although not identical (38). Both ligands have major influences on protein translation via pathways linking receptor activation to AKT and mTOR, and also influence transcriptional programs. Some of the important distinctions in the effects of these ligands may be attributed to differences in early steps in signal transduction pathways, but many of the differences relate rather to the different downstream processes controlled by these networks in different cell types. For example, insulin regulates glycogen storage in liver, but has other roles in other tissues such as mammary gland epithelial cells. Thus, while it is true, as often stated, that insulin has dominantly metabolic effects in contrast to the dominantly mitogenic effects of the IGFs, this generalization is not universal: many cell types can respond mitogenically to insulin via its own receptor at clinically relevant concentrations (for example 19). In contrast, pharmacologic insulin concentrations are commonly used in tissue culture media, as these may act to stimulate cell growth via insulin receptors, IGF-I receptors, or various hybrid receptors.
IGF-I, IGF-II, and insulin obviously share evolutionary ancestry and structural features, but there are obvious distinguishing characteristics. One obvious difference between insulin and the IGFs is the processing of the translated product of the insulin gene to remove c-peptide, a process which does not occur for the IGFs. Other key differences are site of production and control of release. Insulin gene expression is highly restricted to pancreatic cells, where it is tightly regulated according levels of glucose in serum. Furthermore, expression of the gene encoding insulin is not sufficient for the production of bioactive insulin, as specific peptidases are needed to process preproinsulin to the bioactive peptide. Insulin produced by pancreatic -cells acts as a classic hormone, influencing physiology at distant organs, particularly liver, muscle, and fat. Ectopic production of insulin by cell types other than -cells is exceedingly rare.
In contrast, IGF-I and IGF-II have characteristics of tissue growth factors as well as hormones. While most circulating IGFs are produced in the liver, and circulating levels are physiologically regulated, these peptides are also widely expressed in many tissues, where they have important autocrine and paracrine actions.
The gene encoding IGF-II is imprinted (39); thus loss of imprinting represents one of several mechanisms which can result in over-expression of this ligand. There are many reports of over-expression of IGF-I and/or IGF-II by malignant cells, and, in contrast to the well-known example of HER2/neu in an important subset of breast cancers, it appears that the molecular pathology of IGF signalling in neoplasia more often involves inappropriate expression of ligands than derangements of receptors. The bioactivity of IGFs is regulated in a complex fashion by a family of IGF-binding proteins (IGFBPs; 40). These may increase bioactivity by prolonging half life, or decrease bioactivity by competing with receptors for ligands, depending on the physiologic context. IGF-binding proteins are subject to degradation by various proteases, including many secreted by malignant cells; thus secretion of proteolytic enzymes such as prostate-specific antigen (PSA; 41) by transformed cells may increase local IGF bioactivity, contributing to neoplastic behavior.
It is of interest to review the circulating concentrations of insulin ( 0.5 nmol l-1 ), IGF-I ( 20 nmol l-1 ), and IGF-II ( 100 nmol l-1 ) in relation to the above concepts. IGF-II circulates in the highest concentration range of the three ligands, consistent with the fact that access to receptors that transduce its signal is restricted by competition from both the IGF-binding proteins and the IGF-II receptor. Next is IGF-I, which binds to the IGFBPs but not the IGF-II receptor. Finally, insulin, at the lowest concentration range, has a “clear path” to its target receptor, without competition from IGFBPs or the IGF-II receptor.
Circulating levels of IGF-I and IGF-II vary considerably between individuals within a rather broad “normal” range. Twin studies have demonstrated that both genetic and lifestyle factors contribute to this variation (42). Diseases of IGF-I excess (such as acromegaly) and deficit (such as growth-hormone deficiency) are all described, and it is of interest that variation both within the normal range and at pathologic extremes have been associated with cancer risk (1,5,16). Insulin levels of course vary throughout the day to a much greater extent than IGF-I levels, according to food intake. Apart from this level of variation, there are also important differences between individuals over longer timescales; early type II diabetes, obesity, and the “metabolic syndrome” are examples of conditions where insulin levels are higher than normal. In all these cases, the pathophysiology involves attempted compensation by increased insulin secretion to abnormal “insulin resistance” present in classic insulin target organs such as liver, muscle, or adipose tissue. Hyperinsulinemia has been associated with poor cancer prognosis, and there is experimental evidence that this is associated with increased insulin receptor activation in neoplastic tissue (20,21). However, a cause-and-effect relationship is not certain, as insulin resistance is associated with many endocrine abnormalities besides high insulin levels that may influence cancer biology, including, for example, elevated levels of inflammatory cytokines (22). A key issue under study is whether or not insulin resistance is present in insulin-receptor-positive cancers present in patients where systemic insulin resistance is present in classic insulin target organs. If this is not the case, the hyperinsulinemia associated with insulin resistance could stimulate aggressive neoplastic behavior.
Many monoclonal therapeutic antibodies against the IGF-I receptor have been developed. They differ from each other with respect to characteristics such as antibody subclass and serum half life. Many of these antibodies are currently in clinical trials for different potential indications, in many different combinations, and there have been some encouraging results (for example 43,44). On the other hand, figitumumab (Pfizer) is an example of an anti-IGF-I receptor antibody that was taken to Phase III studies in unselected lung cancer patients that revealed not only a lack of benefit, but also toxicity greater than predicted on the basis of Phase II studies (45), and there have been other disappointments (for example 46).
In general, these antibodies were specifically designed to spare insulin receptors, but nevertheless hyperglycemia has been commonly seen, and this is occasionally serious, and associated with other metabolic derangements such as dehydration. Evidence suggests that the metabolic toxicity does not result from cross-reactivity with the insulin receptor, which was successfully avoided by antibody engineering. Rather, it is a consequence of peripheral insulin resistance that arises secondary to the high levels of growth hormone that are associated with administration of these antibodies. The elevations in growth hormone are likely a result of blocked signaling of IGF-I receptors in the hypothalamic–pituitary control systems that regulate serum IGF-I level; perceived IGF-I deficiency leads to increased secretion of growth hormone in order to raise growth-hormone-dependent hepatic IGF-I production. Thus, treatment with anti-IGF-I receptor antibodies is associated with elevations of growth hormone and IGF-I. The high IGF-I levels are without consequence due to receptor blockade, but the high levels of growth hormone lead to peripheral insulin resistance, and occasionally to hyperglycemia and hyperinsulinemia.
Bevacizumab provides a precedent for a therapeutic antibody that targets the ligand (in this example VEGF) rather than the receptor. Several companies are developing high-affinity antibodies that cross-react with the ligands IGF-I and IGF-II (47). These drug candidates are at an earlier stage of development, and clinical toxicity and efficacy data are not yet mature, although pre-clinical data have been regarded as sufficiently positive to justify clinical trials.
Receptor tyrosine-kinase inhibitors
These agents were initially developed as IGF-I receptor-specific inhibitors (48), but it is now recognized that in vivo they broadly inhibit the entire receptor family. In view of this, the theoretical risk for serious metabolic toxicity is relatively high. However, these agents may have advantages over the other approaches in that insulin-receptor-mediated resistance to IGF-I receptor targeting has been described (49), and is less likely with these broader-spectrum inhibitors.
It is of interest that in early clinical trials, the metabolic toxicity of these agents has been less than anticipated, despite the blockade of insulin signaling. This “mystery of safety” is the subject of ongoing research, but some pre-clinical models suggest that pharmacokinetic factors may account for this. There is evidence (23) that the drugs accumulate only to suboptimum therapeutic levels in muscle, which is an important site of insulinmediated glucose uptake. Therefore, severe hyperglycemia may not occur as frequently as was predicted because even at target serum levels, insulin-mediated glucose uptake by muscle takes place. However, other theoretical metabolic hazards must be taken into account with these agents, as there is evidence that their use is associated with hyperinsulinemia, and sudden cessation of therapy or a skipped dose could therefore lead to hypoglycemia.
This agent is commonly used in type II diabetes treatment, where it is lowers both hyperglycemia and hyperinsulinemia. These properties, particularly the latter, have led to proposals that it be evaluated as a drug candidate for treating or preventing cancer (reviewed in 26). Interestingly, these proposed mechanisms do not require an interaction of the drug with neoplastic cells, but only in liver, the key site for its therapeutic effect in diabetes. While retrospective epidemiologic studies suggesting reduced cancer burden among diabetics on metformin as compared to diabetics on other therapies have generated enthusiasm, caveats are that metformin has negligible effects on IGF-I levels, and only lowers insulin levels if they are elevated at baseline. On the other hand, this agent, if it accumulates at sufficient concentration in neoplastic tissue, may inhibit signaling downstream of the insulin/IGF-I receptor family by AMPK-dependent inhibition of mTOR (50). There is no question that metformin has a very favorable safety profile, but there are important gaps in knowledge, including factors influencing metformin accumulation in tumors, and the pros and cons of metformin as compared to other biguanides such as phenformin for potential oncologic indications. Many clinical trials with this off-patent agent are in progress.
This compound has an interesting history and credentials as an IGF-I-receptor inhibitor, although other modes of action have not been excluded with certainty (51–53). Pre-clinical and early clinical results have been favorable, and expansion of the clinical trials program is planned.
Clinical trials to date
More than 100 clinical trials relevant to the hypothesis of insulin and IGFs as therapeutic targets are listed in clinicaltrials.gov, and it is beyond the scope of this review to assess them individually. Although reviews of the status of the trials tend to quickly become outdated, the reader is referred to several recent summaries (reviewed in 54).
Toxicity has been greater than predicted, particularly in large Phase III trials carried out in part in non-specialist centers. Metabolic adverse events of IGF-I receptor targeting agents such as hyperglycemia are treatable if promptly diagnosed, and metformin has been used for this purpose. However, for some oncologists, these adverse events differ from those commonly encountered with approved cancer treatments, so clear algorithms must be provided to enable optimum screening for adverse events, and effective management of these complications.
What can we conclude from clinical trials to date? While there are well-documented anecdotal reports of benefit of IGF-IR targeting agents, particularly in Ewing’s sarcoma, and also some encouraging Phase II results, negative Phase III results clearly must be regarded as more definitive.
However, careful consideration must be given to the generality of conclusions from these trials. One view would be that the negative Phase III data are sufficient to justify abandonment of further investigation of therapeutic targeting of the insulin/IGF-I receptor family for all indications, regardless of targeting strategy. At the opposite extreme, to use the figitumumab example (45), some would limit the conclusion to the specific demonstration that this particular drug candidate has no utility in unselected patients with non-small-cell lung cancer when added to a specific cytotoxic regime, but not extend this conclusion to other possible figitumumab indications, and certainly not to other anti-receptor antibodies or to other targeting strategies, such as small-molecule kinase inhibitors.
The wide spectrum of opinion on this issue is illustrated by the fact that some pharmaceutical companies are closing drug development programs, while others are initiating or continuing trials.
What are the major unanswered questions?
Are there candidate predictive biomarkers that might guide further trials, and is this worth investigating?
There are clear precedents where the use of a predictive biomarker has been essential to define a subset of patients who have tumors for which a particular therapy is applicable, ranging from the classic case of trastuzumab for HER2-positive breast cancer to the more recent example of crizotinib for the small subset of lung cancers that are driven by alk fusion proteins (55). In some cases, the candidate predictive biomarker was obvious, but at the time of initiation of clinical trials of agents that target the insulin/IGF-I receptor family, no validated predictive biomarkers for these drugs had been described. At this time, candidate predictive biomarkers have been identified and some of them are supported by early clinical evidence, but none are validated.
This issue is of central importance in further development of agents that target the insulin/IGF receptor family, now that emerging Phase III data are demonstrating lack of activity in unselected patients, at least for certain indications. It is not necessary that predictive biomarkers perform perfectly, but rather that they allow enrichment of clinical trials of patients more likely to respond, and/or exclude those patients less likely to benefit or more likely to experience toxicity.
Among candidate predictive biomarkers is the pretreatment level of circulating free IGF-I (56). Data to support this candidate comes from evaluation of Phase II rather than Phase III studies, so no definitive conclusions are available, but it is intriguing that there is evidence that confining treatment to subjects with higher free IGF-I levels would reduce toxicity and increase efficacy. The rationale offered for these findings is that tumors that developed in hosts with higher ligand levels were more likely to become dependent on or even addicted to IGF-I receptor activation, and therefore more likely to respond to interruption of signaling.
Additional candidate predictive biomarkers include receptor levels and the presence of autocrine loops. The presence of autocrine loops is relatively easy to assess in tumor specimens, by measuring expression of receptors and ligands. Consideration of this candidate predictive biomarker leads to the question of efficacy of agents for tumors that may be stimulated by circulating ligands from the host (“endocrine” sources), as compared to those that are stimulated by locally produced ligands. Various drug candidates may differ in their efficacy depending on ligand source. For example, higher tissue levels of anti-ligand antibodies may be required to inhibit growth of tumors that have strong autocrine production, as compared to those that rely on circulating ligands. While it is likely that the presence of autocrine loops indicates a degree of dependency of tumors on the signaling pathway, it may be a marker of sensitivity for those agents capable of interrupting such loops, or a marker of resistance for agents that are capable only of attenuating receptor activation related to circulating ligands.
Other candidate predictive biomarkers are under investigation, including, for example, the presence of certain transforming fusion proteins that appear to have a requirement for IGFI-receptor activation (57). In contrast, while not investigated in clinical-trial specimens, it is plausible that the presence of activating mutations downstream of the IGF-I receptor, such as those resulting in constitutive rather than ligand-dependent activation of PI3K, would confer resistance to receptor targeting.
What resistance mechanisms have been proposed?
There are likely many advanced cancers that have evolved to a degree that their behavior is constitutively aggressive and uninfluenced by growth signals. For such neoplasms, targeting the insulin/IGF-I receptor (or any other receptor kinase) will be ineffective. Similarly, some cancers are driven by other receptors (e.g. HER2/neu) to such an extent that insulin/IGF-IR signaling becomes irrelevant, and these would also be predicted to be resistant. Finally, as mentioned above, there may be situations where cancers are dependent on insulin/ IGF-I-receptor activation, but due to processes such insulin-receptor-mediated resistance to IGF-I-receptor targeting or the presence of a strong autocrine loop, particular drugs may not be effective. Many model systems exist that provide examples of cancers that are sensitive to insulin/IGF-I-receptor family targeting, but further research is needed to allow estimates of the percentage of clinical cancers that exhibit this sensitivity. While it is likely that a majority of cancers are indeed resistant, there are examples of successful drug development programs in the era of “personalized medicine” based on sensitivity of only 10% of patient tumors (55).
Are there important differences between the targeting strategies or between drug candidates in a particular class?
As mentioned above, there are several classes of agents that target the insulin/IGF-I-receptor family, and some of these classes (particularly the anti-receptor antibodies) have many individual drug candidates. Most investigators would agree that the various classes of agent that involve different therapeutic strategies are distinct from each other to the extent that toxicity or efficacy data concerning one class will not necessarily apply to the others. However, the significance of differences between agents within a strategic class (e.g. anti-receptor antibodies) is less clear. Many drug candidates were developed for competitive reasons, and although they are clearly non-identical in potentially important areas, including pharmacokinetic profile, it is a matter of controversy if results with any one member of a class are relevant to the other members of the class. This is a matter of practical importance, as it is important to avoid repeating large phase trials of agents that are highly likely to yield results already obtained with similar drug candidates, while at the same time it would be unfortunate if a superior drug candidate was not taken to the clinic on the basis of a prior negative clinical trial experience with an inferior member of a therapeutic class.
Are there rational combination therapies to be explored?
Many targeted therapies are routinely used in combination with other agents, including, as a classic example, trastuzumab. Others, such as crizotinib, are active as single agents. Most trials with agents that target insulin/ IGF-I receptors have involved combinations, but in general these combinations have not been selected on the basis of specific synergy demonstrated preclinically, but rather on a pragmatic approach involving the addition of a drug candidate to a current “standard” therapy that has some activity, but where there is an obvious clinical need to improve efficacy.
Further pre-clinical studies may guide clinical trial design in this area, and offer advantages over a strictly pragmatic approach. For example, synthetic lethality experiments suggest co-targeting partners for agents that inhibit insulin and/or IGFI signaling (58). It is possible that common resistance mechanisms to approved targeted therapies, radiotherapy (59), or cytotoxic agents involve insulin/IGF-IR signaling (60–65), and such information could suggest specific clinical trial designs. Data concerning the roles of the insulin/IGF-I receptor family in mediating resistance to mTOR inhibitors (66) and BRAF inhibitors (67) are of particular interest.
Co-targeting steroid receptors and the insulin/IGF-I receptor family may offer some specific opportunities. For example, recent pre-clinical data indicating that insulin can stimulate local androgen production by prostate cancer cells (68) suggests the possibility of combinations with castration and/or inhibitors of androgen synthesis.
In the search for new cancer therapies, it is well known that investigation of most targets initially considered promising does not justify drug development, that pre-clinical investigation of most drug candidates does not justify clinical evaluation, and that a great majority of drug candidates that are evaluated in clinical trials do not lead to approved indications. It also is clear that among approved drugs, many have therapeutic activity that is clearly documented, but is small in magnitude and/or confined only to small subgroups of patients. Progress by small increments is more common than major leaps forward.
In the case of drug candidates that target the insulin/IGFreceptor family for indications in oncology, initial Phase III trials have been disappointing, and the best-case scenario proposed by optimists 10 years ago, namely broad-spectrum activity for unselected patients with many kinds of advanced cancer, is disproven. However, it remains to be seen if these results represent the beginning of the end for this therapeutic target, or if further studies will reveal specific contexts where targeting this receptor family will lead to clinical benefit. It has been pointed out (69) that, in general, clinical trials designed to detect benefit should take into account the possibility of benefit confined to patient subgroups, and this has not been explained in detail in the case of insulin/ IGF-IR-targeting therapy.
Prominent among current lines of investigation are the search for rational rather than arbitrary drug combinations, the identification of predictive biomarkers, and the pros and cons of each of the various classes of targeting agents, such as anti-receptor antibodies, anti-ligand antibodies, receptor kinase inhibitors, biguanides such as metformin, and others, in terms of safety and efficacy.
- Pollak M. Insulin and insulin-like growth factor signalling in neoplasia. Nature Reviews Cancer 2008;8:915–28.
- Sachdev D, Yee D. Disrupting insulin-like growth factor signaling as a potential cancer therapy. Molecular Cancer Therapeutics 2007;6:1–12.
- Zha J, Lackner MR. Targeting the insulin-like growth factor receptor-1R pathway for cancer therapy. Clinical Cancer Research 2010;16:2512–17.
- Gallagher EJ, LeRoith D. The proliferating role of insulin and insulin-like growth factors in cancer. Trends in Endocrinology and Metabolism 2010;21:610–18.
- Pollak MN, Schernhammer ES, Hankinson SE. Insulin-like growth factors and neoplasia. Nature Reviews Cancer 2004;4:505–18.
- Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 1998;279: 563–6.
- Ma J, Li H, Giovannucci E, et al. Prediagnostic body-mass index, plasma C-peptide concentration, and prostate cancer-specific mortality in men with prostate cancer: a long-term survival analysis. Lancet Oncology 2008;9:1039–47.
- Yang XF, Beamer W, Huynh HT, Pollak M. Reduced growth of human breast cancer xenografts in hosts homozygous for the ‘lit’ mutation. Cancer Research 1996;56:1509–11.
- Sell C, Rubini M, Rubin R, et al. Simian virus 40 large tumor antigen is unable to transform mouse embryonic fibroblasts lacking type 1 insulin-like growth factor receptor. Proceedings of the National Academy of Sciences USA 1993;90:11217–21.
- Ma J, Pollak M, Giovannucci E, et al. Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I, and IGF-binding protein-3. Journal of the National Cancer Institute 1999;91:620–5.
- Rowlands MA, Gunnell D, Harris R, et al. Circulating insulin-like growth factor peptides and prostate cancer risk: a systematic review and meta-analysis. International Journal of Cancer 2009; 124: 2416–29.
- Roddam AW, Allen NE, Appleby P, et al. Insulin-like growth factors, their binding proteins, and prostate cancer risk: analysis of individual patient data from 12 prospective studies. Annals of Internal Medicine 2008;149:461–8.
- Schernhammer ES, Holly JM, Hunter DJ, Pollak MN, Hankinson SE. Insulin-like growth factor-I, its binding proteins (IGFBP-1 and IGFBP-3), and growth hormone and breast cancer risk in The Nurses Health Study II. Endocrine-Related Cancer 2006;13:583–92.
- Hankinson SE, Willett WC, Colditz GA, et al. Circulating concentrations of insulin-like growth factor-I and risk of breast cancer. Lancet 1998;351:1393–6.
- Patel AV, Cheng I, Canzian F, et al. IGF-1, IGFBP-1, and IGFBP-3 polymorphisms predict circulating IGF levels but not breast cancer risk: findings from the Breast and Prostate Cancer Cohort Consortium (BPC3). PLoS One 2008;3:e2578.
- Guevara-Aguirre J, Balasubramanian P, Guevara-Aguirre M, et al. Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans. Science Translational Medicine 2011;3:70ra13.
- Giovannucci E, Harlan DM, Archer MC, et al. Diabetes and cancer: a consensus report. CA, A Cancer Journal for Clinicians 2010;60:207–21.
- Pritchard KI, Shepherd LE, Chapman JW, et al. Randomized trial of tamoxifen versus combined tamoxifen and octreotide LAR therapy in the adjuvant treatment of early-stage breast cancer in postmenopausal women: NCIC CTG MA.14. Journal of Clinical Oncology 2011; 29:3869–76.
- Mashhedi H, Blouin M-J, Zakikhani M, et al. Metformin abolishes increased tumor 18F-2-fluoro-2-deoxy-D-glucose uptake associated with a high-energy diet. Cell Cycle 2011; 10:2770–8.
- Algire C, Amrein L, Bazile M, et al. Diet and tumor LKB1 expression interact to determine sensitivity to anti-neoplastic effects of metformin in vivo. Oncogene 2011;30:1174–82.
- Venkateswaran V, Haddad AQ, Fleshner NE, et al. Association of diet-induced hyperinsulinemia with accelerated growth of prostate cancer (LNCaP) xenografts. Journal of the National Cancer Institute 2007;99:1793–800.
- Park EJ, Lee JH, Yu GY, et al. Dietary and genetic obesity promote liver inflammation and tumorigenesis by enhancing IL-6 and TNF expression. Cell 2010;140:197–208.
- Dool CJ, Mashhedi H, Zakikhani M, et al. IGF-1/insulin receptor kinase inhibition by BMS-536924 is better tolerated than alloxan-induced hypoinsulinemia and more effective than metformin in the treatment of experimental insulin responsive breast cancer. Endocrine-Related Cancer 2011; 18:699–709.
- Seshasai SR, Kaptoge S, Thompson A, et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. New England Journal of Medicine 2011;364:829–41.
- Jee SH, Ohrr H, Sull JW, et al. Fasting serum glucose level and cancer risk in Korean men and women. Journal of the American Medical Association 2005;293: 194–202.
- Pollak M. Metformin and other biguanides in oncology: advancing the research agenda. Cancer Preventiona and Research (Philadelphia) 2010;3:1060–5.
- Wild SH. Diabetes, treatments for diabetes and their effect on cancer incidence and mortality: attempts to disentangle the web of associations. Diabetologia 2011;54:1589–92.
- Horwitz DL, Starr JI, Mako ME, Blackard WG, Rubenstein AH. Proinsulin, insulin, and C-peptide concentrations in human portal and peripheral blood. Journal of Clinical Investigation 1975;55:1278–83.
- Hedman CA, Lindstrom T, Arnqvist HJ. Direct comparison of insulin lispro and aspart shows small differences in plasma insulin profiles after subcutaneous injection in type 1 diabetes. Diabetes Care 2001;24:1120–1.
- Hamaty M, Miller M, Gerstein H, et al. Is insulin exposure associated with higher risk of cancer related hospitalization or death? Analysis of 5 year data from the Accord trial. Proceedings of Annual Meeting of Endocrine Society 2011, Boston. Abstr. 0365.
- Cox M, Gleave M, Zakikhani M, et al. Insulin receptor expression by human prostate cancers. The Prostate 2009;69:33–40.
- Law JH, Habibi G, Hu K, et al. Phosphorylated insulin-like growth factor-1/insulin receptor is present in all breast cancer subtypes and is related to poor survival. Cancer Research 2008;68:10238–46.
- Sprynski AC, Hose D, Kassambara A, et al. Insulin is a potent myeloma cell growth factor through insulin/IGF-1 hybrid receptor activation. Leukemia 2010;24:1940–50.
- Belfiore A, Frasca F, Pandini G, Sciacca L, Vigneri R. Insulin receptor isoforms and insulin receptor/insulin-like growth factor receptor hybrids in physiology and disease. Endocrine Reviews 2009;30:586–623.
- Brown J, Jones EY, Forbes BE. Keeping IGF-II under control: lessons from the IGF-II-IGF2R crystal structure. Trends in Biochemistry Sciences 2009;34:612–19.
- Litzenburger BC, Creighton CJ, Tsimelzon A, et al. High IGF-IR activity in triple-negative breast cancer cell lines and tumorgrafts correlates with sensitivity to anti-IGF-IR therapy. Clinical Cancer Research 2011;17:2314–27.
- Dziadziuszko R, Merrick DT, Witta SE, et al. Insulin-like growth factor receptor 1 (IGF1R) gene copy number is associated with survival in operable non-small-cell lung cancer: a comparison between IGF1R fluorescent in situ hybridization, protein expression, and mRNA expression. Journal of Clinical Oncology 2010;28:2174–80.
- Najjar SM, Blakesley VA, Li CS, et al. Differential phosphorylation of pp120 by insulin and insulin-like growth factor-1 receptors: role for the C-terminal domain of the beta-subunit. Biochemistry 1997;36:6827–34.
- Kaneda A, Feinberg AP. Loss of imprinting of IGF2: a common epigenetic modifier of intestinal tumor risk. Cancer Research 2005;65:11236–40.
- Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins. Endocrine Reviews 2002;23:824–54.
- Cohen P, Graves CB, Peehl DM, et al. Prostate-specific antigen (PSA) is an insulin-like growth factor binding protein-3 protease found in seminal plasma. Journal of Clinical Endocrinology and Metabolism 1992;75: 1046–53.
- Harrela M, Koinstinen H, Kaprio J, et al. Genetic and environmental components of interindividual variation in circulating levels of IGF-I, IGF-II, IGFBP-1, and IGFBP-3. Journal of Clinical Investigation 1996;98:2612–15.
- Kurzrock R, Patnaik A, Aisner J, et al. A Phase I study of weekly R1507, a human monoclonal antibody-insulin-like growth factor-I receptor antagonist, in patients with advanced solid tumors. Clinical Cancer Research 2010;16:2458–65.
- McCaffery I, Tudor Y, Deng H, et al. Effect of baseline (BL) biomarkers on overall survival (OS) in metastatic pancreatic cancer (mPC) patients (pts) treated with ganitumab (GAN; AMG 479) or placebo (P) in combination with gemcitabine (G). Journal of Clinical Oncology 2011; (suppl.; abstr. 4041).
- Jassem J, Langer CJ, Karp DD, et al. Randomized, open label, Phase III trial of figitumumab in combination with paclitaxel and carboplatin versus paclitaxel and carboplatin in patients with non-small cell lung cancer (NSCLC). Journal of Clinical Oncology 2010; (suppl.; abstr. 7500).
- Reidy DL, Vakiani E, Fakih MG, et al. Randomized, Phase II study of the insulin-like growth factor-1 receptor inhibitor IMC-A12, with or without cetuximab, in patients with cetuximab- or panitumumab-refractory metastatic colorectal cancer. Journal of Clinical Oncology 2010;28:4240–6.
- Gao J, Chesebrough JW, Cartlidge SA, et al. Dual IGF-I/II-neutralizing antibody MEDI-573 potently inhibits IGF signaling and tumor growth. Cancer Research 2011;71:1029–40.
- Mitsiades CS, Mitsiades NS, McMullan CJ, et al. Inhibition of the insulin-like growth factor receptor-1 tyrosine kinase activity as a therapeutic strategy for multiple myeloma, other hematologic malignancies, and solid tumors. Cancer Cell 2004;5:221–30.
- Buck E, Gokhale PC, Koujak S, et al. Compensatory insulin receptor (IR) activation on inhibition of insulin-like growth factor-1 receptor (IGF-1R):rationale for cotargeting IGF-1R and IR in cancer. Molecular Cancer Therapeutics 2010;9:2652–64.
- Zakikhani M, Dowling R, Fantus IG, Sonenberg N, Pollak M. Metformin is an AMP kinase-dependent growth inhibitor for breast cancer cells. Cancer Research 2006;66:10269–73.
- Vasilcanu D, Girnita A, Girnita L, et al. The cyclolignan PPP induces activation loop-specific inhibition of tyrosine phosphorylation of the insulin-like growth factor-1 receptor. Link to the phosphatidyl inositol-3 kinase/Akt apoptotic pathway. Oncogene 2004;23:7854–62.
- Klinakis A, Szabolcs M, Chen G, et al. Igf1r as a therapeutic target in a mouse model of basal-like breast cancer. Proceedings of the National Academy of Sciences USA 2009;106:2359–64.
- Girnita A, Girnita L, del Prete F, et al. Cyclolignans as inhibitors of the insulin-like growth factor-1 receptor and malignant cell growth. Cancer Research 2004;64:236–42.
- Gualberto A, Pollak M. Emerging role of insulin-like growth factor receptor inhibitors in oncology: early clinical trial results and future directions. Oncogene 2009;28:3009–21.
- Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. New England Journal of Medicine 2010; 363:1693–1703.
- Gualberto A, Hixon ML, Karp DD, et al. Pre-treatment levels of circulating free IGF-1 identify NSCLC patients who derive clinical benefit from figitumumab. British Journal of Cancer 2011;104:68–74.
- Tognon CE, Somasiri AM, Evdokimova VE, et al. ETV6-NTRK3-mediated breast epithelial cell transformation is blocked by targeting the IGF1R signaling pathway. Cancer Research 2011;71:1060–70.
- Potratz JC, Saunders DN, Wai DH, et al. Synthetic lethality screens reveal RPS6 and MST1R as modifiers of insulin-like growth factor-1 receptor inhibitor activity in childhood sarcomas. Cancer Research 2010;70:8770–81.
- Turner BC, Haffty BC, Narayanan L, et al. Insulin-like growth factor-I receptor overexpression mediates cellular radioresistance and local breast cancer recurrence after lumpectomy and radiation. Cancer Research 1997; 57:3079–83.
- Gooch JL, Van Den Berg CL, Yee D. Insulin-like growth factor-1 rescues breast cancer cells from chemotherapy-induced cell death-proliferative and anti-apoptotic effects. Breast Cancer Research Treat 1999;56:1–10.
- Browne BC, Crown J, Venkatesan N, et al. Inhibition of IGF1R activity enhances response to trastuzumab in HER-2-positive breast cancer cells. Annals of Oncology 2011; 22:68–73.
- Lu Y, Zi X, Zhao Y, Mascarenhas D, Pollak M. Insulin-like growth factor-I receptor signaling and resistance to trastuzumab (herceptin). Journal of the National Cancer Institute 2001;93:1852–7.
- Nahta R, Yuan LX, Zhang B, Kobayashi R, Esteva FJ. Insulin-like growth factor-I receptor/human epidermal growth factor receptor 2 heterodimerization contributes to trastuzumab resistance of breast cancer cells. Cancer Research 2005;65:11118–28.
- Buck E, Eyzaguirre A, Rosenfeld-Franklin M, et al. Feedback mechanisms promote cooperativity for small molecule inhibitors of epidermal and insulin-like growth factor receptors. Cancer Research 2008;68:8322–32.
- Chakravarti A, Loeffler JS, Dyson NJ. Insulin-like growth factor receptor I mediates resistance to anti-epidermal growth factor receptor therapy in primary human glioblastoma cells through continued activation of phosphoinositide 3-kinase signaling.Cancer Research 2002;62:200–7.
- O’Reilly KE, Rojo F, She QB, et al. mTOR inhibition induces upstream receptor tyrosine kinase signaling and activates Akt. Cancer Research 2006;66:1500–8.
- Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K. Cancer Cell 2010;18:683–95.
- Lubik AA, Locke JA, Adomat HH, et al. Insulin directly increases de novo steroidogenesis in prostate cancer cells. Cancer Research 2011; 71:5754–64.
- Betensky RA, Louis DN, Cairncross JG. Influence of unrecognized molecular heterogeneity on randomized clinical trials. Journal of Clinical Oncology 2002;20:2495–9.