Validating immunohistochemical staining for KIT (CD117).

Abstract

There has been rapid evolution in the understanding of the biology of gastrointestinal stromal tumors (GISTs), beginning in 1998 with the first identification of activating mutations in c-kit,1 which encodes a transmembrane receptor tyrosine kinase believed to be critical in the tumorigenesis of the overwhelming majority of GISTs,2 to recognition that immunohistochemical detection of KIT (CD117) could facilitate differential diagnosis of intra-abdominal spindle cell neoplasms,3 to the development of specific inhibitors of tyrosine kinases that already have moved into the clinical arena as therapy for these tumors that previously were untreatable once they relapsed. Currently, immunohistochemical detection of KIT is arguably the most important diagnostic criterion for the diagnosis of GISTs.4 First developed as a small molecule inhibitor of the bcr-abl fusion protein of chronic myelogenous leukemia,5,6 imatinib mesylate (Gleevec, Novartis Pharmaceuticals, East Hanover, NJ) also has been shown to be a selective inhibitor of the tyrosine kinase activity of KIT.7,8 The demonstration of substantial efficacy of imatinib in the treatment of patients with advanced GISTs9 has stimulated widespread clinical interest in detecting KIT expression in tumors other than GISTs. It must be remembered, however, that KIT expression does not equate with functional activation; to date, the only mesenchymal tumors in which c-kit–activating mutations (and resultant constitutively activated kit oncoprotein) have been demonstrated are GISTs.2 Nevertheless, reports of KIT expression in other tumors have perpetuated the hope that inhibitors of KIT tyrosine kinase activity may show therapeutic benefit in patients afflicted with tumors other than GISTs. Immunohistochemical studies of KIT expression using different commercially available antibodies under variable experimental conditions have yielded conflicting results, most strikingly in desmoid fibromatosis,3,10-12 ranging from entirely negative results3 to a study that reported 75% of desmoid tumors positive for KIT.10 In this issue of the Journal, in response to the highly variable findings in previously published reports, Lucas and colleagues13 report a notably thorough evaluation of KIT immunohistochemical analysis with polyclonal antibodies from Santa Cruz Biotechnology (C-19, Santa Cruz, CA) and DAKO (A4502, Carpinteria, CA), the 2 most widely used commercially available antibodies against KIT, in cases of desmoid fibromatosis. The authors evaluated 19 desmoid tumors for KIT expression using 3 dilutions of each antibody, with and without heat-induced epitope retrieval. Careful attention was given to (nonspecific) background staining of stroma and normal structures, and the staining intensity was compared with that of positive control GISTs and internal controls (mast cells) for each experimental condition. The authors found that at low dilutions of both antibodies, a substantial number of desmoid tumors were positive. Interestingly, heat-induced epitope retrieval decreased the number of positive cases with C-19, while the frequency of positive tumors increased with A4502. For both antibodies, however, substantial background stromal staining was noted at low dilutions. At higher dilutions of both antibodies, unwanted background staining was lost, and, with rare exceptions, desmoids were no longer positive. With C-19, GIST control samples and mast cells stained inconsistently at higher dilutions, whereas strong positivity of controls was maintained with A4502.