Author + information
- Received May 14, 2012
- Accepted May 29, 2012
- Published online January 22, 2013.
- ↵⁎Reprint requests and correspondence:
Dr. Thomas Force, Center for Translational Medicine and Cardiology Division, Temple University, School of Medicine, 3500 North Broad Street, MERB Bldg., Room 981, Philadelphia, Pennsylvania 19140
Cancer genomics has focused on the discovery of mutations and chromosomal structural rearrangements that either increase susceptibility to cancer or support the cancer phenotype. Protein kinases are the most frequently mutated genes in the cancer genome, making them attractive therapeutic targets for drug design. However, the use of some of the kinase inhibitors (KIs) has been associated with toxicities to the heart and vasculature, including acute coronary syndromes and heart failure. Herein we discuss the genetic basis of cancer, focusing on mutations in the kinase genome (kinome) that lead to tumorigenesis. This will allow an understanding of the real and potential power of modern cancer therapeutics. The underlying mechanisms that drive the cardiotoxicity of the KIs are also examined. The preclinical models for predicting cardiotoxicity, including induced pluripotent stem cells and zebrafish, are reviewed, with the hope of eventually being able to identify problematic agents before their use in patients. Finally, the use of biomarkers in the clinic is discussed, and newer strategies (i.e., metabolomics and enhanced imaging strategies) that may allow earlier and more accurate detection of cardiotoxicity are reviewed.
The discipline known as “cardio-oncology” or “onco-cardiology” seems to be growing at a rapid pace, driven at least in part by the fact that cancer patients typically have cardiovascular disease and vice versa, demanding a team approach. Another driver of this apparent boom in interest is the increasing awareness of the toxicities, in particular cardiac toxicities, associated with cancer therapeutics. Initially centered on anthracycline- and trastuzumab-induced cardiotoxicity, similar issues are now being faced with some of the so-called “targeted therapeutics,” which largely target protein kinases that are activated or overexpressed, and thus drive growth of various cancers (1,2).
Nearly one-half of the 518 protein kinases encoded by the human genome are expressed from loci associated with specific diseases or regions amplified in human cancers (3). Furthermore, kinases are the most frequently mutated genes in the cancer genome, making them attractive therapeutic targets for drug design. Indeed, these targeted therapeutics have radically altered approaches to the treatment of a number of cancers and now dominate drug discovery and development. Targeted anticancer drugs were initially thought to affect tumors but not normal tissue in which kinases were not constitutively active. Thus, the hope of targeted therapy was one of high efficacy, with minimal adverse effects compared with traditional chemotherapy. However, unexpected reports of cardiotoxicity from approved targeted drugs suggest that these agents are not magic bullets (4). In fact, as long as a kinase maintained expression in the heart, targeting that kinase in cancer could, theoretically, cause cardiotoxicity. Indeed, there are numerous overlapping signaling pathways that drive tumorigenesis but that are also required for cardiomyocyte survival or function (Fig. 1). Cardiotoxicity with kinase inhibitors (KIs) could be said to occur “predictably” if a given molecular target participates in ≥1 basic cardiac function.
The cardio-oncologist also needs to understand cancer and its biology. Central to cancer management is the understanding of cancer as the cumulative phenotypic consequence of acquired genetic and epigenetic alterations in cancer cells. Herein, we also highlight recent advances in cancer genetics and genomics that convey the potential of this discipline to drive personalized cancer medicine. This review also illustrates the importance of defining the biological relevance of genomic data as a key step in realizing the full clinical potential of developing better drugs for cancer management with minimal cardiotoxicity. The rapid emergence of genome technology allows the complete molecular profiling of a patient's tumor. Kinome-wide screening, combined with kinase-targeted libraries of small molecules, and high-throughput structure determination should allow the discovery of selective molecules that will further validate kinases as key targets in oncology (5). This approach should also advance our rather limited understanding of the role that those kinases which are expressed in the human heart play in cardiac biology.
We are entering a new era, one in which personalized medicine in cancer patients is becoming a reality. The success of drug development in cancer, while limiting cardiotoxicity, will require a multidisciplinary approach, with collaborations between oncologists, cardiologists, pharmacologists, and toxicologists, as well as a strong commitment from industry and the regulatory agencies to ongoing translational research. This evolving discipline demands that cardiotoxicity be considered, both in preclinical and clinical scenarios. This issue should be foremost in the mind of investigators during protocol development, enrollment, and follow-up. This should also include postapproval surveillance of patients treated with these drugs for it is then that problems will likely appear when patients with significant comorbidities begin to use these agents. Because small molecule KIs are believed to be the foundation of future cancer treatments, researchers with a particular expertise in this field will be required. Fortunately, there seems to be a growing interest in cardio-oncology at the basic, clinical, and funding agency level (National Cancer Institute and National Heart, Lung, and Blood Institute) and that leads to optimism.
What Do We Know and Do Not Know: Guidelines
Although general guidelines exist for the treatment and follow-up of patients receiving anthracyclines and, to some extent, trastuzumab (Herceptin), that is not the case with the targeted therapeutics (predominantly KIs). This is due in part to the fact that KIs are often unique, with unique targets and, therefore, unique toxicities. To date, we are aware of only 3 attempts to generate guidelines for the care of cancer patients receiving targeted therapeutics. The first was a working group sponsored by the Heart Failure Association of the European Society of Cardiology (6). Although the document focused primarily on anthracyclines and trastuzumab, it was a valiant attempt to try to highlight the gaps in knowledge concerning targeted therapeutics and to begin to address these gaps. Most importantly, the document highlighted how little is understood about the KIs, including the very limited data on their cardiac toxicities. In addition, they noted the need for preclinical screening strategies and also addressed the potential use of biomarkers of injury, specifically troponins and B-type natriuretic peptide, although there were (and continue to be) very little data on the use of these agents in patients being treated with KIs; the exception is some work conducted with trastuzumab (7). This working group was also the first to address the class of agents that are particularly problematic: the antiangiogenesis agents (sunitinib and sorafenib). It was also suggested that registries should be used for potentially problematic agents. Finally, this group formulated several recommendations for patient care as well as guides for the design of future clinical trials. This reasoned and balanced document is very much worth reading.
After the publication of this paper and the clear documentation of cardiotoxicity with some KIs (8,9), the National Cancer Institute convened 2 panels: 1 to address the hypertension that can be profound with antiangiogenesis agents and the other to address cardiotoxicity (10,11). These working groups were specifically charged with developing guidelines for the management of patients who were enrolled in clinical trials sponsored by the National Cancer Institute. Both panels developed general management strategies, although we think it is fair to say that there was serious debate surrounding the cardiotoxicity guidelines. This debate is to be expected when 2 disciplines, both of which want the best for the patient, realize that their strategies are at cross-purposes. This cognitive dissonance is at the core of why cardio-oncology will be such an important discipline going forward. In summary, much remains to be learned about targeted therapeutics and their use, and only when we have a better understanding of them both in cancers and in the heart and vasculature can true guidelines be generated.
The Genetic Basis of Cancer
Cancer is a genetic disease. The cancer cell arises as a clone that expands in an unregulated fashion, driven in part by sequential accumulation of mutations (12,13). Thus, all cancers arise as a result of changes in the DNA sequence of the genomes of cancer cells (14).
The first consistent genetic abnormality associated with human cancer was the translocation that created the Philadelphia chromosome, leading to a fusion of 2 protein kinases, breakpoint cluster region and Abelson leukemia virus tyrosine kinase (ABL), in chronic myeloid leukemia (CML) (15). This led to constitutive activation of the ABL kinase that promoted immortality in myeloid progenitor cells in the bone marrow, which leads to the leukemia. However, because no consistent chromosomal alterations were found in other forms of cancer, it was thought that chromosome alterations were a result of cancer, not a cause. In 1976, however, Stehelin et al. (16) demonstrated that oncogenes (i.e., genes that participate in the initiation of cancer), particularly the tyrosine kinase Schmidt-Ruppin A-2 viral oncogene homolog (SRC), had nontransforming counterparts (proto-oncogenes) that were expressed in normal cells. This work helped to establish the role of genetic alterations in initiating tumorigenesis rather than arising as a result of tumorigenesis (17). Thus, SRC became the first human proto-oncogene to be identified, and many more proto-oncogenes were discovered soon after. Finally, it was demonstrated that introduction of total genomic DNA from human cancers into phenotypically normal (NIH3T3) cells could convert them into cancer cells (18,19).
Somatic Genetic Alterations and Cancer Development
All cancers arise as a result of somatically acquired changes in the DNA (17,20). This does not mean, however, that all of the somatic abnormalities present in a cancer genome are involved in the development of the cancer. Indeed, it is likely that some (or most) make no contribution at all. Thus, the DNA sequence of a cancer cell genome (and of most normal cell genomes) has acquired a set of mutations. These are collectively termed somatic mutations to differentiate them from germ line mutations, which are inherited from parents (12,14). The rate of acquisition and the types of somatic mutations can be increased by exposures that cause DNA damage. These mutations are mitigated by DNA repair processes. If DNA repair fails, somatic mutations will likely increase.
There are 3 categories of cancer genomic aberrations: base mutations, copy number alterations (gain or loss), and translocations/rearrangements (21). These alterations include irreversible changes in the DNA sequence or structure, and in the number of sequences, genes, or chromosomes. These changes can affect hundreds of genes and/or regulatory transcripts (e.g., promoter regions). Epigenetic modifications of DNA or histones by methylation, acetylation, and other mechanisms also became recognized as key mediators of the cancer phenotype (12,14). Collectively, they result in the activation or inhibition of various biological events involved in cancer pathophysiology, including angiogenesis, metastasis, and altered cell growth (22).
Somatic mutations tend to be randomly distributed throughout the genome and may be classified as “driver” or “passenger” mutations (20,22). A driver mutation is causally implicated in oncogenesis because it confers a growth advantage to the cells. Cancer cells expand, invade into surrounding tissue, and typically metastasize. The number of driver mutations in a cancer cell reflects the number of mutated cancer genes and thus drives the cell biological processes required to convert a normal cell into a cancer clone (17,20). The remaining mutations are passengers and do not contribute to cancer development. Passenger mutations are found within cancer genomes because somatic mutations without functional consequences often occur during cell division. A cell that acquires a driver mutation will already have biologically inert somatic mutations within its genome (13). Thus, the catalog of somatic mutations in the genome of a cancer cell represents genomic changes that usually accumulate over several decades (12,23,24). There are usually between 1,000 and 10,000 somatic mutations in the genome of adult cancers (12,20,25). Catalogue of Somatic Mutations in Cancer is currently the most comprehensive resource for information on somatic mutations in human cancer. Almost 4,800 genes and 250,000 tumors have been examined, resulting in >50,000 mutations being available for investigation (24).
Genome-Wide Approaches for Cancer Gene Discovery
The emergence of the human genome sequence advanced the study of cancer genomes in many ways. It led to a fundamental shift toward global views of genomes and transcriptomes in human biology and disease. This progress was made possible by increasingly powerful technologies. By the late 1990s, microarray applications and high-throughput sequencing began to explore entire cancer genomes (12,21,26). These advances, together with advances in bioinformatics, enabled a link between tumor genomic alterations and critical functional roles in cancers (21). These ongoing global genome characterizations are revolutionizing cancer biology and management. The goal of the Cancer Genome Atlas is to catalog and discover major cancer-causing genome alterations in large cohorts of human tumors. The International Cancer Genome Consortium has similar goals, with 39 teams planning to study 18,000 tumor genomes in 50 different cancers (25). It is strikingly apparent that the researchers in cardiovascular disease need similar initiatives.
Second-generation DNA sequencing technologies and their application to cancer have accelerated the pace of investigations of cancer genomes (27,28), enabling the complete (and efficient) sequencing of entire genomes (27). One key advantage is that second-generation sequencing offers structural information never before available, allowing assessment of chromosomal rearrangements in cancer (23,27). These technologies have allowed sequencing of >2,000 individual cancers (12).
The Cancer Kinome and its “Druggability”
The kinome of an organism is the set of protein kinases in its genome. The term was first used in 2002 in a report analyzing the 518 human protein kinases (1). Kinases are enzymes that catalyze phosphorylation of amino acids and are grouped into 3 major families: kinases that phosphorylate the amino acids serine and threonine, those that phosphorylate tyrosine, and some that can phosphorylate both. Approximately 90 of the 518 kinases in the human kinome are tyrosine kinases (3). Since the discovery of the first oncogene, vSrc, in the 1970s and its identification as an enzyme with tyrosine kinase activity (29,30), dysfunctional signaling by mutated or overexpressed kinases has been intimately linked to cancer (31). There are numerous oncogenes, many of which are protein kinases (17). Protein kinases may act as tumor suppressors or proto-oncogenes in normal, healthy cells, but mutations in these may lead to tumorigenesis through inactivation of a tumor suppressor or constitutive activation of a proto-oncogene, transforming it into an oncogene. The approval of the first KI, imatinib, for the treatment of CML in 2001 and the success of antibody-based drugs that targeted the epidermal growth factor receptor (HER2) in breast cancer (trastuzumab) and colon cancer (cetuximab) heralded the current period of KI development (5). More than 10,000 patent applications for KIs having been filed since 2001 in the United States alone. To date, approximately 80 KIs have progressed to the point of clinical evaluation and 13 KIs have been approved for cancer treatment (32) (Table 1).
Genetic Alterations in the Kinome and Cancer
Kinases and their direct regulators are the most frequently mutated oncogenes and tumor suppressors (17,25,33). The MoKCa database (Mutations of Kinases in Cancer) was developed to annotate, and predict, the phenotypic consequences of mutations in protein kinases implicated in cancer (34). Recent resequencing of the kinome in cancer cell lines has revealed that most somatic mutations are likely to be passengers (20). Although difficult, differentiating passengers from drivers is critical for understanding the molecular mechanisms responsible for tumor initiation and progression and also provides prognostic information as well as targets for therapeutic intervention (35).
Targeting the Cancer Kinome: KIs as Anticancer Agents
Many human malignancies are linked with activated kinases or inactivated phosphatases, and one third of the targets under investigation by pharmaceutical companies are kinases (36). Two classes of therapeutics have been developed to target the cancer kinome: humanized monoclonal antibodies (mAbs) and KIs. The mAbs bind cancer cell–specific antigens, commonly to the extracellular portion of receptor tyrosine kinases, thereby inhibiting tyrosine kinase activation. The binding of mAbs to the extracellular domain of the receptor tyrosine kinase can block ligand binding to the receptor and inhibit subsequent dimerization and activation of the kinase (37,38). Trastuzumab (Herceptin) binds the HER2 receptor, but other mAbs bind the growth factors that activate the receptors. For example, bevacizumab (Avastin) targets vascular endothelial growth factor (VEGF) A, preventing it from interacting with the VEGF receptor and leading to inhibition of tumor angiogenesis (37).
The dependency of certain cancers on 1 or a few genes for maintenance of the malignant phenotype, termed oncogenes addiction, provides a strong rationale for molecular targeting in cancer therapy. KIs have revolutionized the treatment of select cancers, such as CML and gastrointestinal stromal tumor, which are driven by a single oncogenic kinase. For these conditions, in most cases, KIs have led to multiyear increases in survival (39). One other issue involves acquired resistance to KIs. This resistance is typically due to de novo mutations in the kinase that reduce binding affinity of the drug to the kinase. In some cases, it has been possible to address this problem with redesign of the inhibitor; for example, the second-generation breakpoint cluster region–ABL KIs dasatinib and nilotinib.
Critical to the activity of the kinase is adenosine triphosphate (ATP) binding in the cleft within the kinase “hinge” (32). The ATP cleft contains structural elements responsible for the catalytic activity of the kinase. All kinases have a conserved A loop, and this region is the target of most KIs in use or in development (33).
On the basis of the mode of binding to kinases, KIs can be classified into 4 different types (Table 2). Type I inhibitors constitute the majority of ATP-competitive inhibitors and recognize the active conformation of the kinase (32,36). Because the structure of the ATP pocket is highly conserved across the human kinome, it is relatively easy to make an inhibitor that targets the ATP pocket of the kinase of interest. Due to this, most KIs discovered to date are ATP-competitive inhibitors. However, it is not surprising that lack of selectivity is an issue with most of the inhibitors in this class (38). By contrast, the type II inhibitors recognize the inactive conformation of the kinase and still have contacts to the ATP pocket, but they also interact with a site adjacent to the pocket. This action allows enhanced selectivity and binding to the inactive confirmation. These KIs are typically more potent than the type I inhibitors (32,36). Type III inhibitors (non–ATP-competitive inhibitors) bind to regions remote from the ATP pocket. Inhibitors in this class tend to exhibit the highest degree of selectivity because they exploit binding sites and regulatory mechanisms unique to a particular kinase. However, lack of structural information for these regions represents a major hurdle in designing inhibitors (32,40). A fourth class (covalent inhibitors) is capable of forming an irreversible, covalent bond to the kinase-active site (40). Pre-clinical studies have shown that irreversible inhibitors inhibit EGFR signaling even in gefitinib-resistant cell lines (harboring a T790M mutation). Therefore, tyrosine kinase inhibitors may overcome resistance associated with mutations such as T790M (41). A number of these new-generation TKIs have been developed (41). Despite the large number of kinases that could be targeted by this approach, there are concerns about the potential for toxicity of covalent inhibitors as a result of permanent modification of unanticipated targets or even permanent systemic inhibition of anticipated targets.
Molecular Mechanisms of Cardiotoxicity
The first report of cardiotoxicity with tyrosine kinase inhibitors was a case series of 10 patients who developed congestive heart failure while receiving imatinib (8). Subsequently, much more serious toxicity was identified with other molecularly targeted therapies. There are 2 general types of cardiotoxicity: on-target and off-target.
With on-target toxicity, the kinase that is targeted in the cancer also provides an important function in the heart and/or vasculature (38). Inhibiting this kinase leads to adverse consequences in the heart. An example of this is sorafenib. In addition to inhibiting several growth factor receptors, sorafenib inhibits rapidly accelerated fibrosarcoma (RAF)-1 and BRAF. RAF family kinases function in the pro-survival ERK (or mitogen-activated protein kinase) cascade. The importance of RAF-1 in the heart has been demonstrated in mouse models in which conditional cardiac-specific deletion of RAF-1 resulted in left ventricular (LV) dilation and reduced contractile function (42). Escudier et al. (43) demonstrated the efficacy of sorafenib in advanced clear-cell carcinoma (kidney). Cardiotoxicity was observed in the form of cardiac ischemia or infarction (reported to occur in 3% of patients) (38). We and others have observed cases of probable sorafenib-induced LV dysfunction and heart failure, but we do not know the rate of these events because careful studies examining this question have not been done.
Approaches to deal with unavoidable on-target toxicity have been proposed and include targeted delivery of drug specifically to the cancer, sparing normal tissue, or inhibiting cell death pathways in the heart that are activated by a compound but that are not necessary for tumor cell death. For the latter, JNK inhibition has been proposed as a strategy to limit imatinib-induced cardiomyocyte death without reducing antitumor efficacy (44). Another approach could be to identify and target unique kinases from the cancer kinome that are either not expressed in the heart or are not required for normal heart function. So far, target selection in drug discovery has been strongly biased toward previously validated targets. It seems that the majority of agents in development target kinases for which approved drugs are already available. Recently, novel kinase targets that are uncharacterized have been identified via genome sequencing and ribonucleic acid interference screens (5).
Off-target toxicity is directly related to nonselectivity. In this scenario, inhibition of a target not intended to be inhibited by a KI is responsible for the cardiotoxicity. Almost all of the approved KIs bind to multiple kinases, although with different affinities. Cardiotoxicity with KIs is directly correlated with a lack of target specificity (45). Often, the kinase mediating the toxicity will not be known, and the complexity of identifying the key target, inhibition of which leads to cardiotoxicity, can be very challenging. This is the main reason that we believe enhanced selectivity of KIs should be a goal.
An example of the problem with poor selectivity is sunitinib-induced cardiotoxicity. Sunitinib was designed as a multikinase inhibitor that targeted VEGF receptor 1, platelet-derived growth factor, colony-stimulating factor 1 receptor, and several others. However, it seems that, at therapeutic plasma levels, sunitinib would inhibit approximately 90 kinases (46). The reported incidence of cardiotoxicity (primarily LV dysfunction) with sunitinib was low in the early clinical trials. However, a more thorough retrospective review of data reported in Chu et al. (51) in a study of 75 patients with gastrointestinal stromal tumor found an 8% incidence of heart failure and a 28% incidence of LV ejection fraction decrease of >10%.
Toxicity of sunitinib is likely due to both on- and off-target effects. Off-target cardiotoxicity could be avoided (or reduced) by improving selectivity and/or redesigning the drug so that it no longer inhibits the crucial kinase (assuming the kinase does not play a key part in cancer progression). Obviously, off-target toxicity will be much more frequent with multitargeted KIs and combination therapy.
Preclinical Detection of Cardiotoxicity
Our limited understanding of the role various kinases play in the heart makes it impossible, at least at this point in time, to predict which KIs will have a high probability of associated cardiotoxicity. Therefore, it is essential that valid preclinical models be developed. That said, it is our sense that most models, including rodents, are relatively insensitive, due at least in part to the lack of comorbidities such as hypertension and coronary artery disease in these animals (47). Cancer patients typically have cardiovascular comorbidities, and inhibition of protective signaling pathways by KIs, which would be well tolerated in the rodent, might not be in patients who have cardiovascular disease. Underscoring the importance of comorbidities was a study, albeit small, which identified the presence of coronary artery disease as the best predictor of cardiotoxicity with the KI sunitinib (9). Similarly, sunitinib can cause striking hypertension in patients yet we saw only mild hypertension in mice. Strategies to stress the heart, such as thoracic aortic constriction or infusion of pressors, will induce hypertension, but these are costly and time-consuming procedures that are technically very demanding (at least with thoracic aortic constriction).
As a result, investigators have searched for alternatives. Zebrafish have been successfully used to screen for drugs that might prolong QT intervals. Cheng et al. (48) used zebrafish in an attempt to identify agents that might be cardiotoxic; they were able to detect cardiotoxicity with sunitinib and sorafenib based on quantification of contractile dysfunction. They also used a transgenic fish that expresses a red fluorescent protein in cardiomyocytes and were able to determine that both drugs led to a reduction in total cardiomyocyte number. More surprisingly, they were able to interrogate signaling pathways that were disrupted in the drug-treated fish and to identify a cardioprotective pathway that preserved ventricular function. The next step would be to use the so-called Casper fish that, unlike wild-type zebrafish, remains translucent throughout life. Thus, LV function can be examined over time in the adult fish, eliminating drug-induced toxicity as a result of developmental effects.
Recent advances in the development and production of stem cell–derived cardiomyocytes has allowed an assessment of this model to study cardiac arrhythmia and prediction of cardiotoxicity (49). Human stem cell–derived cardiomyocytes exhibit consistent rhythmic contractions, and prolonged culture increasingly promotes the expression of adult cardiac proteins (50,51). In addition, studies using these stem cell–derived cardiomyocytes helped investigate the role for sunitinib in inducing cardiac arrhythmia and cardiotoxicity, as well as the potential role for 2 kinases, AMPK and RSK, which have been speculated to modulates these toxicities (52). Although generic cytotoxicity observed in cells in vitro has been questioned as a reliable predictor of organ toxicity (53), the persistent rhythmic beating of stem cell–derived cardiomyocytes provides a unique mechanistic advantage in studying changes in cardiomyocyte function, contraction, and energy homeostasis that occur during KI-mediated cardiotoxicity.
Clinical Detection of Cardiotoxicity: Biomarkers and Imaging
Although significant strides have been made in identifying biomarkers, including troponin I (the most studied), that predict cardiotoxicity with anthracyclines, much less is known about the use of biomarkers in the setting of KI treatment. The only work that we are aware of is with the mAb trastuzumab (7). Thus, it is entirely unclear whether these “traditional” biomarkers will be effective in KI-induced cardiotoxicity. We suspect that better biomarkers will be needed to detect toxicity with the KIs, and we propose that more sensitive metabolomic strategies should be developed along the lines of the work done by Gerszten et al. (54).
As for imaging, encouraging results in patients treated with anthracyclines and trastuzumab have been reported with a combination of longitudinal strain and high-sensitivity troponin I (55). It remains to be seen whether this approach will be useful in the “KI-alone” treated patients.
Conclusions and Future Directions
Targeted therapeutics, particularly KIs, have radically altered cancer treatment. However, given the central role played by kinases in the heart and other organs (leading to on-target toxicity with KIs), the relatively poor selectivity of the current (and future) crop of KIs, leading to off-target effects, the paucity of verified preclinical models to predict toxicity, the uncertainties as to how effective traditional biomarkers and imaging will be in identifying cardiotoxicity with these agents, our poor understanding of the role played by the majority of kinases expressed in the heart, and the number of new KIs in development, we are heading into uncharted waters.
The authors apologize to those whose original work could not be cited owing to space limitations.
This work was supported by grant HL061688 and HL091799 from the U.S. National Heart, Lung, and Blood Institute (to Dr. Force). Dr. Force received research funding, as well as consultancy fees, from GlaxoSmithKline. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- Abelson leukemia virus tyrosine kinase
- adenosine triphosphate
- chronic myeloid leukemia
- kinase inhibitors
- left ventricular
- monoclonal antibodies
- rapidly accelerated fibrosarcoma
- Schmidt-Ruppin A-2 viral oncogene homolog
- vascular endothelial growth factor
- Received May 14, 2012.
- Accepted May 29, 2012.
- American College of Cardiology Foundation
- Manning G.,
- Whyte D.B.,
- Martinez R.,
- Hunter T.,
- Sudarsanam S.
- Maitland M.L.,
- Bakris G.L.,
- Black H.R.,
- et al.
- Stratton M.R.
- Nowell P.C.,
- Hungerford D.A.
- Krontiris T.G.,
- Cooper G.M.
- Macconaill L.E.,
- Garraway L.A.
- Chin L.,
- Hahn W.C.,
- Getz G.,
- Meyerson M.
- Forbes S.A.,
- Bhamra G.,
- Bamford S.,
- et al.
- Wood L.D.,
- Parsons D.W.,
- Jones S.,
- et al.
- Collett M.S.,
- Erikson R.L.
- Richardson C.J.,
- Gao Q.,
- Mitsopoulous C.,
- Zvelebil M.,
- Pearl L.H.,
- Pearl F.M.
- Chen M.H.,
- Kerkela R.,
- Force T.
- Cheng H.,
- Force T.
- Minotti G.,
- Salvatorelli E.,
- Menna P.
- French K.J.,
- Coatney R.W.,
- Renninger J.P.,
- et al.
- Cheng H.,
- Kari G.,
- Dicker A.P.,
- Rodeck U.,
- Koch W.J.,
- Force T.
- Pekkanen-Mattila M.,
- Chapman H.,
- Kerkela E.,
- et al.
- Lin Z.,
- Will Y.
- Gerszten R.E.,
- Asnani A.,
- Carr S.A.
- What Do We Know and Do Not Know: Guidelines
- The Genetic Basis of Cancer
- Somatic Genetic Alterations and Cancer Development
- Genome-Wide Approaches for Cancer Gene Discovery
- The Cancer Kinome and its “Druggability”
- Genetic Alterations in the Kinome and Cancer
- Targeting the Cancer Kinome: KIs as Anticancer Agents
- Molecular Mechanisms of Cardiotoxicity
- Preclinical Detection of Cardiotoxicity
- Clinical Detection of Cardiotoxicity: Biomarkers and Imaging
- Conclusions and Future Directions