Anatomic Pathology / KIT IN SOFT TISSUE SARCOMAS Immunohistochemical Staining for KIT (CD117) in Soft Tissue Sarcomas Is Very Limited in Distribution Jason L. Hornick, MD, PhD, and Christopher D.M. Fletcher, MD, FRCPath Key Words: KIT; CD117; GIST; STI571; Sarcoma; Immunohistochemistry Abstract We performed immunohistochemical analysis for KIT in 365 soft tissue sarcomas. Most tumors evaluated were completely negative for KIT, including all cases of leiomyosarcoma, rhabdomyosarcoma, myxofibrosarcoma, liposarcoma, solitary fibrous tumor, synovial sarcoma, dermatofibrosarcoma protuberans, schwannoma, malignant peripheral nerve sheath tumor, clear cell sarcoma, low-grade endometrial stromal sarcoma, and follicular dendritic cell sarcoma. Tumors showing occasional immunoreactivity for KIT included extraskeletal myxoid chondrosarcoma (2/20), Ewing sarcoma/malignant primitive peripheral neuroectodermal tumor (4/20), melanotic schwannoma (3/5), metastatic melanoma (4/20), and angiosarcoma (5/20). In most cases, staining for KIT was focal. Rare tumor cells showing KIT positivity were identified in a small number of other tumors. This study demonstrates very limited expression of KIT in soft tissue tumors other than gastrointestinal stromal tumors and underscores the discriminatory value of KIT immunohistochemical analysis for differential diagnosis. As some of these findings differ markedly from previous reports, it is evident again that variations in immunohistochemical technique can lead to major discrepancies in positive staining. Since treatment eligibility for selective tyrosine kinase inhibitors such as STI571 hinges on positive immunostaining, standardization and reproducibility of meaningful results are critically important. c-kit encodes a membrane receptor tyrosine kinase that is expressed in gastrointestinal stromal tumors (GISTs) but not in smooth muscle or neural tumors of the gastrointestinal tract. 1-3 Immunohistochemical staining for KIT (CD117) is therefore increasingly regarded as invaluable in the differential diagnosis of intra-abdominal mesenchymal tumors. c-kit also is expressed in limited normal tissue cell types, including mast cells, germ cells, melanocytes, breast epithelium, and interstitial cells of Cajal, 4-7 the pacemaker cells of the gastrointestinal tract from which lineage GISTs are thought to originate. 2,8 Mutations in c-kit leading to the constitutive activation of the tyrosine kinase are believed to have a role in the tumorigenesis of the overwhelming majority of GISTs. 9-13 STI571, a small molecule inhibitor of the bcr-abl fusion protein of chronic myelogenous leukemia, 14,15 also has been shown to be a selective inhibitor of the tyrosine kinase activity of KIT. 16 A recent report demonstrated the clinical potential of STI571 in a patient with metastatic GIST. 17 Results of phase 1 and 2 clinical trials establishing the safety and efficacy of STI571 in patients with recurrent or metastatic GISTs were presented recently. 18,19 In the phase 2 study, 19 (54%) of 35 evaluable patients showed partial responses and 12 (34%) showed stable disease. 18 The promise of this novel therapeutic strategy for GISTs has stimulated considerable clinical (and indeed patient) interest in the possible expression of KIT in other tumors, particularly of the mesenchymal type. There have been several reports of KIT immunostaining in a limited number of soft tissue tumors other than GISTs, including metastatic melanoma, clear cell sarcoma, angiosarcoma, extraskeletal Ewing sarcoma, and desmoid fibromatosis, at times with conflicting results. 1,20-22 In the present study, we sought to undertake a comprehensive examination 188 Am J Clin Pathol 2002;117:188-193 American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE Table 1 Summary of KIT Immunostaining in Soft Tissue Tumors Total No. No. (%) of Tumor of Cases Positive Cases Comments Leiomyosarcoma 40 0 (0) Rhabdomyosarcoma 25 0 (0) Myxofibrosarcoma 20 0 (0) Low-grade fibromyxoid sarcoma 10 1 (10) Very rare cells positive Myxoid/round cell liposarcoma 10 0 (0) Dedifferentiated liposarcoma 10 0 (0) Solitary fibrous tumor 20 0 (0) Synovial sarcoma 20 0 (0) Dermatofibrosarcoma protuberans 30 0 (0) Desmoid fibromatosis 20 1 (5) Focal weak positivity Extraskeletal myxoid chondrosarcoma 20 2 (10) 1 case with >50% of cells positive; 1 case with focal weak positivity Ewing sarcoma/malignant primitive 20 4 (20) 1 case with >50% of cells positive; 3 cases with focal positivity peripheral neuroectodermal tumor Schwannoma 20 0 (0) Melanotic schwannoma 5 3 (60) 2 cases with >50% of cells positive; 1 case with focal positivity Perineurioma 10 2 (20) 1 case with >50% of cells positive; 1 case with rare cells positive Malignant peripheral nerve sheath tumor 20 0 (0) Metastatic melanoma 20 4 (20) 2 cases with focal positivity; 1 case with focal weak positivity; 1 case with rare cells positive Clear cell sarcoma 10 0 (0) Angiosarcoma 20 5 (25) 3 cases with focal positivity; 2 cases with focal weak positivity Low-grade endometrial stromal sarcoma 10 0 (0) Follicular dendritic cell sarcoma 5 0 (0) of KIT immunohistochemical staining in a large number of soft tissue tumors, both to clarify the discriminant value of KIT immunostaining for differential diagnosis and to examine whether other mesenchymal malignant neoplasms might be candidates for therapeutic strategies involving KIT tyrosine kinase inhibitors such as STI571. We focused principally on sarcomas and other aggressive mesenchymal lesions because of the new therapeutic relevance of KIT expression, but we also included certain benign or locally aggressive spindle cell lesions because they may enter the differential diagnosis of GIST. Materials and Methods Cases were retrieved from the consultation files of one of the authors (C.D.M.F.) or the routine surgical pathology files of Brigham and Women s Hospital, Boston, MA (10 cases) or St Thomas s Hospital, London, England (3 cases). All tumors studied had been formalin-fixed. All diagnoses were based on standard and widely accepted criteria. 23,24 Immunohistochemical staining for KIT (CD117) was performed using a 1:250 dilution of the rabbit polyclonal antibody A4502 (DAKO, Carpinteria, CA) with the EnVision detection system (DAKO). No antigen retrieval method was used. In cases of melanotic schwannoma, the LSAB detection system conjugated with alkaline phosphatase (DAKO) was used to yield a red reaction product. Appropriate positive and negative controls were used throughout. Results We examined 365 soft tissue tumors for KIT expression by immunohistochemical staining Table 1. The presence of mast cells in virtually all cases served as an internal control. There was no tumor staining for KIT in 40 cases of leiomyosarcoma, including 10 cases each of uterine, cutaneous, vascular, and soft tissue tumors. KIT immunostaining also was completely negative in 25 cases of rhabdomyosarcoma, including 10 cases of alveolar type, 10 cases of embryonal or spindle cell type, and 5 cases of pleomorphic type. In addition, 20 cases of myxofibrosarcoma were negative for KIT. One case of 10 low-grade fibromyxoid sarcomas showed very rare tumor cells positive for KIT. All liposarcomas evaluated were negative for KIT, including 10 cases each of myxoid/round cell liposarcoma and dedifferentiated liposarcoma. A prominent mast cell infiltrate showing anticipated KIT positivity was seen in several cases of solitary fibrous tumor (SFT) and synovial sarcoma Image 1 ; however, no staining of tumor cells was observed in 20 cases each of either synovial sarcoma or SFT. Rare stromal endothelial cells showed KIT staining in a case of SFT. Thirty cases of dermatofibrosarcoma protuberans were negative for KIT, including 15 cases of the higher grade fibrosarcomatous variant. One case of extra-abdominal desmoid fibromatosis showed focal weak immunostaining for KIT, whereas the remaining 19 cases of fibromatosis, both intra-abdominal and extra-abdominal, were completely negative. American Society for Clinical Pathology Am J Clin Pathol 2002;117:188-193 189
Hornick and Fletcher / KIT IN SOFT TISSUE SARCOMAS Image 1 A prominent mast cell infiltrate positive for KIT in a case of monophasic synovial sarcoma (immunohistochemical stain, 200). Image 2 Diffuse immunoreactivity for KIT in a case of extraskeletal myxoid chondrosarcoma (immunohistochemical stain, 200). Image 3 A case of Ewing sarcoma/malignant primitive peripheral neuroectodermal tumor showing positivity for KIT (immunohistochemical stain, 200). Image 4 Cytoplasmic staining for KIT in a case of melanotic schwannoma (immunohistochemical stain with alkaline phosphatase, red reaction product, 400). Among 20 extraskeletal myxoid chondrosarcomas, 1 case showed diffuse membranous staining for KIT Image 2, while a second case demonstrated focal weak positivity. Four of 20 cases of Ewing sarcoma/malignant primitive peripheral neuroectodermal tumor (MPNET) were positive for KIT Image 3 ; in 3 of these cases, staining was focal. Whereas all 20 schwannomas examined were negative for KIT, including 10 cases of cellular schwannoma and 4 cases of gastrointestinal schwannoma, 3 of 5 melanotic schwannomas showed positive staining for KIT Image 4. Two of 10 soft tissue perineuriomas were positive for KIT, with 1 case showing diffuse cytoplasmic staining; in the second case, only rare tumor cells were positive. There was no tumor staining for KIT in 20 cases of malignant peripheral nerve sheath tumor. Focal membranous tumor cell staining for KIT was seen in 4 of 20 cases of metastatic melanoma Image 5. All 10 clear cell sarcomas (malignant melanomas of soft tissue) examined were completely negative for KIT. Of 20 cases of angiosarcoma, 5 showed focal staining for KIT Image 6. 190 Am J Clin Pathol 2002;117:188-193 American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE Image 5 Focal membranous staining for KIT in a case of metastatic melanoma (immunohistochemical stain, 400). Image 6 A case of epithelioid angiosarcoma showing focal KIT positivity (immunohistochemical stain, 400). No staining for KIT was seen in either 10 low-grade endometrial stromal sarcomas or 5 follicular dendritic cell sarcomas. Discussion In this study, we found immunohistochemical staining for KIT in a very limited number of soft tissue tumors. In the majority of positive cases, KIT staining was focal. In contrast, GISTs typically show diffuse and generally strong staining for KIT, usually in a cytoplasmic distribution with or without dot-like accentuation. We used the rabbit polyclonal antibody A4502 (DAKO), which has shown consistent performance with low background in our hands. More than 35 consecutive GISTs found to overexpress KIT by immunohistochemical analysis in our laboratory using this antibody have been confirmed to express KIT by immunoblotting (J.A. Fletcher, MD, verbal communication, 2001), and, thus, we have gained confidence with this antibody in more than 400 GISTs that we have examined during the past 3 years (unpublished data). We have been unable to achieve reproducible or consistently interpretable immunohistochemical results with commercially available KIT antibodies used in some other laboratories, although we are aware that others have had more success. 1 However, for the purposes of this study, we selected the A4502 polyclonal antibody based not only on our personal experience but also because it seems to be the most widely used KIT antibody and because it is the antibody specified in the ongoing large cooperative clinical trials of STI571. Therefore, in the interests of practicality and to reflect the majority view of daily clinical practice, this seemed to us to be the most realistic choice. The precise staining protocols recommended in the various trials of STI571 have varied somewhat with regard to antigen retrieval. However, it has been our consistent personal experience with this particular antibody (as with a few others) that staining intensity is actually reduced after heat-induced antigen retrieval; thus, the method used in our laboratory was optimized without this step. Our positive results are concordant with those reported previously for several tumors, although we found fewer KITpositive cases. We observed focal KIT immunoreactivity in 25% of angiosarcomas, while in a previous report, more than half of angiosarcomas were positive for KIT, with half of those cases showing reactivity in more than 50% of tumor cells. 20 We also confirmed focal KIT immunopositivity in metastatic melanoma (20% of cases), which has been described at a somewhat higher frequency by others. 1,21 Another group examining KIT expression in benign and malignant melanocytic lesions found decreasing KIT staining with melanoma progression and the lowest numbers of positive cases in metastatic tumors. 25 Studies in vitro and in mouse models have shown that enforced KIT expression in melanoma cell lines decreases both tumorigenicity and metastases. 26,27 Since loss of KIT seems to be associated with malignancy in melanocytic neoplasia, tyrosine kinase inhibitors are unlikely to be of therapeutic benefit in metastatic melanoma. Interestingly, we found no KIT-positive cases of clear cell sarcoma, although a previous study reported tumor staining for KIT in nearly 50% of such cases. 1 These discrepancies might be attributable to the different commercially available American Society for Clinical Pathology Am J Clin Pathol 2002;117:188-193 191
Hornick and Fletcher / KIT IN SOFT TISSUE SARCOMAS polyclonal antibody used in these other series, which, in our personal experience, is more often associated with high background (and, hence, false-positive) staining. We found KIT immunostaining in 20% of Ewing sarcoma/mpnet cases, which again is approximately half the frequency reported in a recent series. 21 The ligand for KIT, stem cell factor, has previously been shown to protect MPNET and Ewing sarcoma cell lines from apoptosis in vitro. 28 A recent study in a mouse model, however, reported a decrease in the metastatic rate of Ewing sarcoma cells following pretreatment with stem cell factor. 29 These preliminary experiments suggest that KIT may have a role in the malignant phenotype of these tumors. Whether KIT inhibitors might provide therapeutic benefit to the subset of patients with Ewing sarcoma/mpnet whose tumors show immunoreactivity for KIT remains to be determined. Despite a previous report of several cases of dermatofibrosarcoma protuberans that showed focal reactivity for KIT, 1 in our hands, no cases of dermatofibrosarcoma protuberans were positive for KIT, including higher grade fibrosarcomatous cases. Of 20 cases of desmoid fibromatosis, we found focal weak KIT immunostaining in only 1 case. Our results differ strikingly from a recent report that described KIT staining in 75% of intra-abdominal desmoid tumors. 22 Interestingly, we used the same antibody as these authors; however, we used no pretreatment, whereas they performed heat-induced antigen retrieval for KIT immunostaining. The pretreatment conditions used by these authors may have caused nonspecific tumor cell staining. In a subsequent report, 14 cases of fibromatosis purportedly all showed positive KIT staining in tumor cells using the identical antibody with or without heat-induced antigen retrieval. 30 These findings are difficult to reconcile with our data, particularly since through daily clinical practice outside the present study, we have stained at least 20 other cases with similarly negative results. Furthermore, kit-activating mutations and KIT expression by immunoblotting have never been identified in a desmoid fibromatosis (J.A. Fletcher, MD, verbal communication, 2001). It is likely that technical differences in experimental protocol led to this apparent nonspecific staining in fibromatosis. We identified several tumors that have not previously been reported to show KIT immunoreactivity. Among nerve sheath tumors, a subset of melanotic schwannomas and perineuriomas were positive for KIT, although we found no cases of conventional schwannoma or malignant peripheral nerve sheath tumor showing KIT immunostaining. Finally, occasional cases of extraskeletal myxoid chondrosarcoma demonstrated KIT positivity. The significance, if any, of these findings awaits further elucidation through the study of additional cases and molecular genetic analysis. Importantly, we have not seen KIT positivity (in the present study or in a large number of tumors studied in daily clinical practice) in leiomyosarcomas, by far the most frequent differential diagnosis of GIST. For all tumor types examined in the present study, the KIT-positive tumor cells were generally a small minority (<5% in most cases) and, hence, unlikely to be of therapeutic relevance. Because this was a retrospective study, in none of the positive cases did we have suitable material for c-kit sequencing. However, in other cases of Ewing sarcoma and desmoid fibromatosis that have been sequenced in our laboratory, none have shown kit mutation (J.A. Fletcher, MD, verbal communication, 2001). We found that immunostaining for KIT in soft tissue tumors other than GISTs is very limited. As our results differ substantially from those reported by some other laboratories, it is evident that choice of antibody and experimental conditions have a striking effect on the apparent immunoreactivity of tumors. Since treatment eligibility for tyrosine kinase inhibitors such as STI571 hinges on the results of KIT immunohistochemical staining results, reproducibility and the elimination of nonspecific staining are critical. Given the indisputable and widely recognized interlaboratory variability in immunohistochemical staining results in at least a small proportion of human tumors examined in routine clinical practice, the ideal route to achieve consistency is unclear. In our opinion, the notion of a federally approved or mandated antibody or technique is both untenable and unrealistic and, as exemplified by tests for HER2/neu expression, runs the risk of causing confusion with no discernible benefit. In view of the relative ease with which c-kit mutational status can be determined nowadays, molecular genetic corroboration through demonstration of kit-activating mutations in these other tumor types showing occasional KIT positivity would seem desirable before pathologists inadvertently raise false hopes of treatment sensitivity among clinicians and patients. From the Department of Pathology, Brigham and Women s Hospital and Harvard Medical School, Boston, MA. Address reprint requests to Dr Fletcher: Dept of Pathology, Brigham and Women s Hospital, 75 Francis St, Boston, MA 02115. Acknowledgment: We are very grateful to Catherine Quigley for expert technical assistance and supervision. References 1. Sarlomo-Rikala M, Kovatich AJ, Barusevicius A, et al. CD117: a sensitive marker for gastrointestinal stromal tumors that is more specific than CD34. Mod Pathol. 1998;11:728-734. 2. Kindblom L-G, Remotti HE, Aldenborg F, et al. Gastrointestinal pacemaker cell tumor (GIPACT): gastrointestinal stromal tumors show phenotypic characteristics of the interstitial cells of Cajal. Am J Pathol. 1998;152:1259-1269. 192 Am J Clin Pathol 2002;117:188-193 American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE 3. Miettinen M, Sarlomo-Rikala M, Sobin LH, et al. Gastrointestinal stromal tumors and leiomyosarcomas in the colon: a clinicopathologic, immunohistochemical, and molecular genetic study of 44 cases. Am J Surg Pathol. 2000;24:1339-1352. 4. Maeda H, Yamagata A, Nishikawa S, et al. Requirement of c- kit for development of intestinal pacemaker system. Development. 1992;116:369-375. 5. Tsuura Y, Hiraki H, Watanabe K, et al. Preferential localization of c-kit product in tissue mast cells, basal cells of skin, epithelial cells of breast, small cell lung carcinoma and seminoma/dysgerminoma in human: immunohistochemical study on formalin-fixed, paraffin-embedded tissues. Virchows Arch. 1994;424:135-141. 6. Huizinga JD, Thuneberg L, Kluppel M, et al. W/kit gene required for interstitial cells of Cajal and for intestinal pacemaker activity. Nature. 1995;373:347-349. 7. Arber DA, Tamayo R, Weiss LM. Paraffin section detection of the c-kit gene product (CD117) in human tissues: value in the diagnosis of mast cell disorders. Hum Pathol. 1998;29:498-504. 8. Sircar K, Hewlett BR, Huizinga JD, et al. Interstitial cells of Cajal as precursors of gastrointestinal stromal tumors. Am J Surg Pathol. 1999;23:377-389. 9. Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science. 1998;279:577-580. 10. Moskaluk CA, Tian Q, Marshall CR, et al. Mutations of c-kit JM domain are found in a minority of human gastrointestinal stromal tumors. Oncogene. 1999;18:1897-1902. 11. Lasota J, Jasinski M, Sarlomo-Rikala M, et al. Mutations in exon 11 of c-kit occur preferentially in malignant versus benign gastrointestinal stromal tumors and do not occur in leiomyomas or leiomyosarcomas. Am J Pathol. 1999;154:53-60. 12. Lux ML, Rubin BP, Biase TL, et al. KIT extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol. 2000;156:791-795. 13. Lasota J, Wozniak A, Sarlomo-Rikala M, et al. Mutations in exons 9 and 13 of KIT gene are rare events in gastrointestinal stromal tumors: a study of 200 cases. Am J Pathol. 2000;157:1091-1095. 14. Druker BJ, Tamura S, Buchdunger E, et al. Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med. 1996;2:561-566. 15. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med. 2001;344:1031-1037. 16. Heinrich MC, Griffith DJ, Druker BJ, et al. Inhibition of c-kit receptor tyrosine kinase activity by STI 571, a selective tyrosine kinase inhibitor. Blood. 2000;96:925-932. 17. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med. 2001;344:1052-1056. 18. Blanke CD, von Mehren M, Joensuu H, et al. Evaluation of the safety and efficacy of an oral molecularly-targeted therapy, STI571, in patients with unresectable or metastatic gastrointestinal stromal tumors (GISTs) expressing c-kit (CD117) [abstract]. Presented at the 37th Annual Meeting of the American Society of Clinical Oncology; May 12-15, 2001; San Francisco, CA. 19. van Oosterom AT, Judson I, Verweij J, et al. STI571, an active drug in metastatic gastrointestinal stromal tumors (GIST), an EORTC phase I study [abstract]. Presented at the 37th Annual Meeting of the American Society of Clinical Oncology; May 12-15, 2001; San Francisco, CA. 20. Miettinen M, Sarlomo-Rikala M, Lasota J. KIT expression in angiosarcomas and fetal endothelial cells: lack of mutations of exon 11 and exon 17 of c-kit. Mod Pathol. 2000;13:536-541. 21. Miettinen M, Sobin LH, Sarlomo-Rikala M. Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with a reference to CD117 (KIT). Mod Pathol. 2000;13:1134-1142. 22. Yantiss RK, Spiro IJ, Compton CC, et al. Gastrointestinal stromal tumor versus intra-abdominal fibromatosis of the bowel wall. Am J Surg Pathol. 2000;24:947-957. 23. Weiss SW, Goldblum JR. Enzinger and Weiss s Soft Tissue Tumors. 4th ed. St Louis, MO: Mosby; 2001. 24. Fletcher CDM. Diagnostic Histopathology of Tumors. 2nd ed. London, England: Churchill Livingstone; 2000. 25. Montone KT, Belle PV, Elenitsas R, et al. Proto-oncogene c- kit expression in malignant melanoma: protein loss with tumor progression. Mod Pathol. 1997;10:939-944. 26. Huang S, Luca M, Gutman M, et al. Enforced c-kit expression renders highly metastatic human melanoma cells susceptible to stem cell factor induced apoptosis and inhibits their tumorigenic and metastatic potential. Oncogene. 1996;13:2339-2347. 27. Huang S, Jean D, Luca M, et al. Loss of AP-2 results in downregulation of c-kit and enhancement of melanoma tumorigenicity and metastasis. EMBO J. 1998;17:4358-4369. 28. Ricotti E, Fagioli F, Garelli E, et al. c-kit is expressed in soft tissue sarcoma of neuroectodermic origin and its ligand prevents apoptosis of neoplastic cells. Blood. 1998;91:2397-2405. 29. Landuzzi L, De Giovanni C, Nicoletti G, et al. The metastatic ability of Ewing s sarcoma cells is modulated by stem cell factor and by its receptor c-kit. Am J Pathol. 2000;157:2123-2131. 30. Miettinen M. Are desmoid tumors kit positive [letter]? Am J Surg Pathol. 2001;25:549-550. American Society for Clinical Pathology Am J Clin Pathol 2002;117:188-193 193