Tyrosine kinase enzymes are integral to many processes involved in intracellular signalling. They transduce messages from outside the cell to the cell's nucleus. Many tyrosine kinase interactions are known to influence the growth signalling pathways of the cell; the potential result of aberrant tyrosine kinase signalling is the uncontrolled proliferation of the affected cells. In effect, this is cancer.

With an understanding of the molecular biology of disease we can attribute certain medical phenotypes to specific molecular changes. Molecular changes can result in the altered chemistry of that particular molecule. Since all molecules have a three-dimensional shape, it is possible to use computer software to develop a range of potential drug molecules designed to specifically interact only with the altered target molecule. This interaction can effectively negate the effect of the abnormality, potentially restoring the 'wild type' phenotype.

In order to use tyrosine kinase inhibitors (TKIs) successfully, we need to know the target against which we are using the agent. Over the course of the last few decades, cancer science has probed the molecular biology of cancer so that we now understand some of the molecular changes behind the progression of some forms of cancer. The condition for which the first TKI was developed was human chronic myelogenous leukaemia (CML). This condition is defined by the presence of a chromosomal translocation resulting in the creation of a functional tyrosine kinase protein which drives uncontrolled cell proliferation. Since the chromosomal translocation is a defining characteristic of the disease, the probability of a patient exhibiting the molecular target is very high and, as a result, the proportion of patients achieving favourable outcomes was similarly very high. Since the development of imatinib, the first licensed TKI, other products have reached clinical trials but few have actually been awarded licences. There are two reasons for this. As a rule there is far greater heterogeneity between tumours at a molecular level, than we see with CML. Secondly, despite the intention that TKIs exhibit remarkable specificity, they do interact with non-target tyrosine kinase molecules; the impact of these off-target interactions is variable but can be catastrophic.

Imatinib was licensed in 2001 for the treatment of CML. The target for the new drug molecule was the active site of the fusion protein, bcr-abl. As already stated, this protein was a constitutively active growth-signalling molecule. The shape of the active site of the bcr-abl fusion protein was, not surprisingly, similar to those of other tyrosine kinase molecules so it did not take too long to learn that there were other molecular interactions which could have significant consequences. One such molecular interaction was noted between imatinib and a normal tyrosine kinase molecule, c-kit. Normal c-kit is involved in the transduction of signals involved in gastrointestinal motility and in the coordination of inflammatory responses, among many other functions. Interference with normal c-kit function could be seen to be the cause of adverse events associated with the use of imatinib. Medical science being what it is though, it was quickly realised that c-kit was also aberrantly expressed in a different tumour altogether, gastrointestinal stromal tumour (GIST). Like CML, GIST is defined by the presence of expression of a specific molecule, in this case c-kit on the cell surface. By targeting the normal c-kit molecule, successful management of GISTs has been achieved.

The discovery that imatinib can be used in the treatment of GISTs illustrates a point at the heart of this presentation, namely that the specificity of TKIs is imperfect; therefore beneficial interactions can be found which allow the successful management of other histologically unrelated tumours. However, the role of imatinib in GIST management was not found by mere coincidence; it did require an intelligent process of scientific discovery to understand the off-target interactions between imatinib and c-kit and the knowledge that c-kit signalling was integral to the survival of GIST cells.

Knowledge of the molecular pathway instrumental to the growth of a specific tumour creates an opportunity to generate a drug molecule that targets this specific process.

The first veterinary TKI to be awarded a licence was masitinib (Masivet, AB Science), initially licensed for the treatment of grade 2 or 3 cutaneous mast cell tumours exhibiting a mutation of the c-kit growth factor receptor gene. The terms of the licence reflected the wisdom that interference with the appropriate target was critical to the success of the therapy. For the first time in history, the veterinary profession was being asked to account for molecular biology in therapeutic decision making. In the short time since masitinib was made available, it has been recognised that efficacy is not limited to cases with known c-kit mutation only; we currently speculate that there is some efficacy achieved through interactions with normal c-kit expressed in mast cell tumours lacking c-kit mutation though the drug definitely performs less well in this group of patients than the group exhibiting mutation.

The second licensed veterinary TKI, toceranib (Palladia, Pfizer), has a very different spectrum of tyrosine kinase inhibition to masitinib. When canine cutaneous mast cell tumours are treated with toceranib, a greater proportion of cases not exhibiting c-kit mutation respond to therapy than we see with masitinib. The explanation for this is that toceranib targets other pathways by which means cell death can arise. The positive impact of this reduced target specificity then is clearly a broader spectrum of tumours demonstrating a measurable tumour response. However, the price paid is the broader range of cellular processes targeted, and the increased risk of morbidity or death as an unwanted consequence.

In medical decision making, we select therapeutic agents based upon a number of considerations. These include the diagnosis itself, the safety of the agent in question, the probability that the agent will have a favourable impact in the context of the diagnosed cause of illness and non-medical issues such as the cost of therapy, ease of administration and the requirement for and extent of travel and monitoring. Undoubtedly, the medical elements of the decision-making process take primacy. Therefore, in order to make rational prescribing decisions, we need to understand the diagnosis, the medicine and the interactions that define the positive and negative impacts of medicine administration.

In cancer, we are used to treating tumours largely on the basis of their histological diagnosis. In some tumours we add an extra layer of complexity to the decision-making pathway; for example, in the case of mast cell tumours, we refine our management decisions according to the histological grade of the tumour. It is pertinent to note, however, that histological grade is actually simply a proxy measure. Histological grade has been shown to be something that allows the prediction of tumour behaviour with a degree of reproducibility. Tumour behaviour is a reflection of tumour biology and tumour biology is what we treat with our surgery and medicines.

We have lived and worked in an era of cancer diagnosis by histology and treatment by agents that exhibit reasonably uniform toxic effects. Cytotoxic agents primarily achieve toxicity through interference with the process of mitosis. For this reason the side-effect profile of conventional cytotoxic agents is predictable and largely limited to tissues which constantly undergo self-renewal. Tyrosine kinase signalling is integral to almost every function of every living cell. The spectrum of cellular processes potentially affected by the on-target and off-target activity of TKIs is absolute. The interaction profiles for individual TKIs will differ substantially.

Current drug evaluation strategies and regulatory processes do not allow for the identification of all potential interactions between a drug molecule and vital biochemical processes. Ultimately, the true picture of how individual TKIs interact with tumours and patients will only emerge with clinical usage. With the advent of TKIs comes a great opportunity but also a new responsibility. It is incumbent on those of us who choose to use TKIs to ensure that we do so wisely. We must practise good pharmacovigilance and we must work together to best learn how to apply these agents to the greatest good.

References

1.  Druker BJ, Talpaz M, et al. Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. New England Journal of Medicine 2001;344:1031–1037.

2.  Hahn KA, Ogilvie G, et al. Masitinib is safe and effective for the treatment of canine mast cell tumors. Journal of Veterinary Internal Medicine 2008;22:1301–1309.

3.  Hahn KA, Legendre AM, et al. Evaluation of 12- and 24-month survival rates after treatment with masitinib in dogs with nonresectable mast cell tumors. American Journal of Veterinary Research 2010;71:1354–1361.

4.  Letard S, Yang Y, et al. Gain-of-function mutations in the extracellular domain of KIT are common in canine mast cell tumors. Molecular Cancer Research 2008;6:1137–1145.

5.  London CA, Malpas PB, et al. Multi-center, placebo-controlled, double-blind, randomized study of oral toceranib phosphate (SU11654), a receptor tyrosine kinase inhibitor, for the treatment of dogs with recurrent (either local or distant) mast cell tumor following surgical excision. Clinical Cancer Research 2009;15:3856–3865.