Introduction

Prior to the publication of an article in Veterinary Pathology in 2003,¹ gastrointestinal stromal tumour was a term used to indicate a range of mesenchymal tumours of the gastrointestinal tract. Efforts were made to characterise tumours better, but these were confusing times. Immunohistochemistry was able to indicate myogenic and neurogenic differentiation; the failure of tumours to adhere to strict immunohistochemical guidelines also yielded so-called anaplastic and mixed histotypes. However, immunohistochemical efforts to categorise gastrointestinal mesenchymal tumours did not adequately discriminate between types and consequently, little evidence of significant differences between tumours emerged. In human pathology, expression of the cell surface growth factor receptor, c-kit, best distinguished leiomyosarcoma from true gastrointestinal stromal tumour (GIST). Other c-kit-positive cancer entities were easily distinguished on routine H&E evaluations; they included unusual cases of melanoma and extraskeletal Ewing's sarcoma which would never be confused with GIST. In 2003, Frost and Miettinen described leiomyosarcoma and GIST in dogs and provided a framework for their discrimination.¹

Gastrointestinal stromal tumours arise from the neoplastic transformation of the intestinal pacemaker cells, or the "intestinal cells of Cajal." The intestinal cells of Cajal are c-kit positive and indeed, c-kit dependent. Withdrawal of c-kit stimulation leads to differentiation into normal smooth muscle cells. In human patients, 95% of GISTs are c-kit positive. Of the remainder, 35% have activating mutations of the related cell surface growth factor receptor, platelet-derived growth factor receptor alpha (PDGFRA). c-kit is therefore a defining moiety in intestinal tumours of mesenchymal histology.

Of all GISTs studied in human patients, 85% of c-kit-positive tumours have activating c-kit or PDGFRA mutations. In both instances, these mutated growth factor receptors activate the same downstream signalling pathways. These pathways activate molecular cascades responsible for the growth and mitosis of these cells. Autonomous growth follows, which of course underpins the development of neoplasia. Recent work has demonstrated that the activating mutations responsible for the development of human GISTs are mirrored in canine patients.²

Interestingly, in an article published in 2000,³ the authors stated, "Specific identification of GIST may become clinically important if therapies targeting the KIT tyrosine kinase activation become available." One wonders how much these authors showed foresight and how much they showed inside knowledge. Only one year later, imatinib was licensed for treatment of chronic myelogenous leukaemia in people. However, work was already underway to investigate its possible application in GISTs.

Imatinib was designed, by computer, to specifically interact with and negate the growth-promoting effects of a chimaeric protein, bcr-abl. The activity of this protein led to uncontrolled cell proliferation and the development of chronic myelogenous leukaemia. bcr-abl is a fusion protein that arises secondary to a chromosomal translocation and ligation of two-part genes into a single fusion gene. The drug molecule imatinib sits in the active site of the functional part of the bcr-abl fusion protein and prevents it from phosphorylating its downstream target. The active site of the gene is a tyrosine kinase enzyme. There are many other related tyrosine kinase enzyme-based growth factor receptors, including the cell surface receptors c-kit and PDGFRA.

Imatinib, and other related tyrosine kinase inhibitors, successfully inhibit the action of mutated c-kit. 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 signaling was integral to the survival of GIST cells.

When tyrosine kinase inhibitors are used in GIST management, appropriate clinical expectations must be defined. Early experiences with imatinib led to marked improvements in well-being for patients, but doctors would withdraw treatment despite this because their training instructed them that patients whose tumours did not reduce in size by 30–50% were not responding to treatment. Multiple letters and articles followed reporting the experiences of individual patients or patient groups. Work followed in which the standards by which treatment was regarded successful were revised.⁴

Optimal therapy for canine GISTs is surgical removal. The incidence of regional or distant metastasis is reported at 28–36%.^5,6 However, these data come from studies of clinical cases without necropsy data, so the true incidence of metastasis may be much higher. Metastasis does appear to be predictable, employing a tumour grade scheme that is already in use in human medical practice.⁶ Among the simplest indicators of malignancy are tumour location (large intestinal has a more favourable prognosis than small intestinal), tumour size (larger is worse), and mitotic index (higher is poorer).

Until the advent of the tyrosine kinase inhibitors, patients considered to be at high risk of GIST recurrence or metastasis would have been treated with adjuvant doxorubicin or other equally poorly substantiated cytotoxic therapies. However, with the licensing of two veterinary tyrosine kinase inhibitors with proven efficacy against mutated c-kit, there is reason to be optimistic about the treatment possibilities.

At present, very small numbers of cases have been treated. The details of cases managed under my supervision will be presented. This anecdotal evidence will hopefully help delegates make good decisions for the treatment of their own cases in the future. Gastrointestinal stromal tumour serves as an excellent model of how new therapies can be rationally applied in a context which differs from their intended or licensed use.

References

1.  Frost D, Lasota J, Miettinen M. Gastrointestinal stromal tumors and leiomyomas in the dog: a histopathologic, immunohistochemical, and molecular genetic study of 50 cases. Veterinary Pathology. 2003;40:42–54.

2.  Gregory-Bryson E, Bartlett E, Kiupel M, Hayes S, Yuzbasiyan-Gurkan V. Canine and human gastrointestinal stromal tumors display similar mutations in c-KIT exon 11. BMC Cancer. 2010;10:559.

3.  Miettinen M, Sobin LH, and Sarlomo-Rikala M. Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with a reference to CD117 (KIT). Modern Pathology. 2000;13:1134–1142.

4.  Benjamin RS, Choi H, Macapinlac HA, Burgess MA, Patel SR, Chen LL, Podoloff DA, Charnsangavej C. We should desist using RECIST, at least in GIST. Journal of Clinical Oncology. 2007;25:1760–1764.

5.  Hayes S, Yuzbasiyan-Gurkan V, Gregory-Bryson E, Kiupel M. (2013) Classification of canine nonangiogenic, nonlymphogenic, gastrointestinal sarcomas based on microscopic, immunohistochemical and molecular characteristics. Veterinary Pathology. 2013;50:799–788.

6.  Hanazono K, Fukumoto S, Hirayama K, Takashima K, Yamane Y, Natsuhori M, Kadosawa T, Uchide T. Predicting metastatic potential of gastrointestinal stromal tumors in dogs by ultrasonography. The Journal of Veterinary Medical Science. 2012;74:1477–1482.