RESEARCH ARTICLE

DOG1 for the Diagnosis of Gastrointestinal Stromal Tumor (GIST): Comparison Between 2 Different Antibodies Lisandro F. Lopes, MD, PhD,* Robert B. West, MD, PhD,w Livia M. Bacchi, MS,z Matt van de Rijn, MD, PhD,w and Carlos E. Bacchi, MD, PhD*

Abstract: Gastrointestinal stromal tumor (GIST) is the most common mesenchymal neoplasm of the gastrointestinal tract. Discovered on GIST-1 (DOG1) is a recently described protein expressed in GISTs irrespective of mutation status. The aim of this study was to investigate the immunohistochemical expression of DOG1 using 2 different monoclonal antibodies (DOG1.1 and the commercially available K9 antibody) in 668 GIST cases and to compare the results with the expression of KIT. DOG1 and KIT expression also were studied in most human normal tissues and several nonmesenchymal and mesenchymal tumors other than GIST. KIT was expressed in 643 (96.3%) GISTs. DOG1.1 and K9 were positive in 538 (80.5%) and 642 (96.1%) GIST cases, respectively. In 25 (3.7%) KIT-negative GIST cases, DOG1 was expressed in 5 (20.0%) and 19 (76.0%) using DOG1.1 and K9 antibodies, respectively. Only 0.9% of GISTs were negative for KIT, DOG1.1, and K9. Most normal human tissues did not reveal KIT and DOG1 expression. DOG1.1 was positive in only 2 of 57 synovial sarcomas and 1 of 61 soft tissue leiomyosarcomas. K9 was positive in 5 of 57 synovial sarcomas, 1 of 14 angiosarcomas, 1 of 61 soft tissue leiomyosarcomas, 3 of 4 adenoid cystic carcinomas of the head and neck, and in myoepithelial cells of 9 of 11 fibroadenomas of the breast. In conclusion, the commercially available K9 is of great utility for the diagnosis of most KIT-negative GISTs, and the combination of both KIT and K9 antibody in a panel of immunohistochemistry can define the diagnosis of GIST in more than 99% of cases. Key Words: DOG1, gastrointestinal stromal tumor, GIST, immunohistochemistry, KIT

receptor a (PDGFRA) activating mutations]. KIT and PDGFRA encode for similarly named receptors (KIT and PDGFRA) that can be inactivated by tyrosine kinase inhibitors, such as imatinib and sunitinib. On account of the availability of specific treatment, the accurate diagnosis of GIST is of great importance.1–8 KIT immunohistochemical expression is observed in most GISTs, whereas other mesenchymal neoplasms that enter in the differential diagnosis of GIST are generally negative; however, up to 5% of GISTs are KIT-negative.9,10 ‘‘Discovered on GIST-1’’ (DOG1) is a protein encoded by TMEM16A (also known under several names that include TMEM16, FLJ10261, ANO1, ORAOV2, and TAOS2), a gene found to be highly expressed in GISTs by gene-expression profiling.11 In these tumors, DOG1 is strongly expressed on the surface of the neoplastic cells irrespective of mutation status, being rarely expressed in other soft tissue tumors, as demonstrated by earlier studies that used a polyclonal antiserum and the monoclonal antibody DOG1.1.11–13 A new monoclonal antibody, clone K9, has become commercially available recently and was analyzed on a large series of GISTs.14 The purpose of this study was to compare the immunohistochemical expression of DOG1 as detected by 2 different monoclonal antibodies, DOG1.1 and K9, and contrast this with the expression of KIT on a series of 668 GIST cases that include 25 KIT-negative GISTs. The expression of KIT and DOG1 in most normal human tissues and several tumors other than GISTs (mesenchymal and nonmesenchymal) was also investigated.

(Appl Immunohistochem Mol Morphol 2010;18:333–337)

MATERIALS AND METHODS

G

astrointestinal stromal tumor (GIST) is the most common mesenchymal neoplasm of the gastrointestinal tract, and it is characterized by distinct immunohistochemical (KIT/CD117 expression) and molecular genetic findings [KIT or platelet-derived growth factor

Received for publication December 3, 2009; accepted January 4, 2010. From the *Consultoria em Patologia, Botucatu; zSchool of Medicine, University of Sa˜o Paulo, Sa˜o Paulo, SP, Brazil; and wDepartment of Pathology, Stanford University Medical Center, Stanford, CA. Reprints: Carlos E. Bacchi, MD, PhD, Consultoria em Patologia, Rua Major Leonidas Cardoso, 739 Botucatu, Sa˜o Paulo, SP 18602-010, Brazil (e-mail: [email protected]). Copyright r 2010 by Lippincott Williams & Wilkins

Appl Immunohistochem Mol Morphol



Case Material The study group included 668 cases of GIST diagnosed between 1997 and 2008 at Consultoria em Patologia, a reference consultation service in anatomic pathology located in Botucatu, Sao Paulo State, Brazil. All cases had clinical and morphologic features of GIST, and the diagnosis was supported by earlier immunohistochemical and/or molecular genetic studies. Normal human tissues including adrenal, pancreas, small and large intestine, stomach, skeletal muscle, spleen, thyroid, lymph node, thymus, bone marrow, kidney, liver, uterine cervix, esophagus, skin, peripheral nerve, placenta, lung, prostate, pituitary gland, cartilaginous tissue, brain, breast, salivary gland, parathyroid, ovary, appendix, bladder,

Volume 18, Number 4, July 2010

www.appliedimmunohist.com |

333

Lopes et al

Appl Immunohistochem Mol Morphol

epididymis, cardiac muscle, umbilical cord, endometrium, and myometrium also were included in this study as well as tumors other than GIST.



Volume 18, Number 4, July 2010

Tissue Microarray Construction and Immunohistochemistry Tissue microarray blocks were constructed using a tissue arrayer (Beecher Instruments, Sun Prairie, WI). Each GIST case was represented by 3 tumor cores of 0.6 mm taken from the original paraffin block of each case. Appropriate positive and negative controls were included. Cores (0.6 mm) of normal human tissues and tumors other than GIST also were distributed over different array blocks. Serial sections of 3 mm were cut from the tissue array blocks and used for immunohistochemical analysis. The immunohistochemical expression of KIT (CD117) and DOG1 using 2 different antibodies (DOG1.1 monoclonal antibody and the commercially available K9 monoclonal antibody as specified in Table 1) was investigated. The antigens, primary antibodies (clones), dilutions, antigen retrieval methods, detection systems, and sources are listed in Table 1. Diaminobenzidine was the chromogen. Staining by KIT (CD117), DOG1.1, and K9 were considered positive if cytoplasmic/membranous expression was observed in more than 10% of the neoplastic cells in at least 1 of the 3 cores of each case, irrespective of its intensity. The majority of GIST cases showed similar staining (positive or negative) in all 3 cores with any of the 3 antibodies.

RESULTS KIT and DOG1 Expression in GIST KIT (CD117) expression was observed in 643 (96.3%) cases. DOG1 expression using DOG1.1 monoclonal antibody was detected in 538 (80.5%) GISTs and in 642 (96.1%) cases using the commercially available K9 monoclonal antibody. Moreover, the intensity of expression was, in general, much stronger with K9 than DOG1.1 antibody (Fig. 1). In the 25 (3.7%) GISTs that were KITnegative, DOG1.1 was positive in 5 (20.0%) of the cases, whereas K9 was positive in 19 (76.0%) cases (Table 2). Only 6 (0.9%) GISTs were negative for KIT and DOG1 using DOG1.1 and K9 antibodies. These 6 cases had clinical and morphologic findings indicative of GIST.

KIT and DOG1 Expression in Normal Human Tissues Most normal human tissues were negative for KIT and DOG1 (DOG1.1 and K9 antibodies). KIT was expressed only in the testis (seminiferous tubules) and seminal vesicles. DOG1.1 was positive in seminal vesicles only. K9 was positive in seminal vesicles and in gallbladder mucosal epithelial cells.

KIT and DOG1 Expression in Tumors Other Than GIST FIGURE 1. DOG1 expression in a case of GIST. A, Spindle cell GIST (hematoxylin and eosin). B, K9 antibody positivity. C, DOG1.1 antibody positivity. Note the stronger intensity of DOG1 expression using K9 in contrast to DOG1.1 antibody in the same case of GIST. GIST indicates gastrointestinal stromal tumor.

334 | www.appliedimmunohist.com

KIT was expressed in 2 of 5 (40.0%) papillary thyroid carcinomas, 1 of 4 (25.0%) adenoid cystic carcinomas of the head and neck, 1 of 11 (9.1%) fibroadenomas of the breast, 3 of 15 (20.0%) melanomas, 4 of 17 (23.5%) Ewing sarcoma/primitive neuroectodermal r

2010 Lippincott Williams & Wilkins

Appl Immunohistochem Mol Morphol



Volume 18, Number 4, July 2010

Comparison of DOG1 Antibodies

TABLE 1. Panel of Antibodies Used in this Study Antigen

Clone

Dilution

DOG1

DOG1.1

1 : 3200

HIER (steamer)

DOG1

K9

1 : 200

HIER (pressure cooker)

Polyclonal

1 : 150

HIER (microwave)

KIT (CD117)

Antigen Retrieval

Detection System Novolink Polymer, Leica Microsystems, Newcastle Upon Tyne, UK Novolink Polymer, Leica Microsystems, Newcastle Upon Tyne, UK Envision Mouse, Dako, Carpinteria, CA

Source Matt van de Rijn (Stanford University) Leica Microsystems, Newcastle Upon Tyne, UK Dako, Carpinteria, CA

DOG1 indicates discovered on GIST-1; GIST, gastrointestinal stromal tumor; HIER, heat-induced epitope retrieval.

tumor cases, 1 of 14 (7.1%) angiosarcomas, 1 of 30 (3.4%) soft tissue schwannomas, 1 of 57 (1.7%) synovial sarcomas, and 1 of 72 (1.4%) soft tissue undifferentiated pleomorphic sarcomas. DOG1 as detected by the DOG1.1 antibody was only expressed in 2 of 57 (3.5%) synovial sarcomas and 1 of 61 (1.6%) soft tissue leiomyosarcomas. The K9 antibody was positive in 5 of 57 (8.8%) synovial sarcomas, 1 of 14 (7.1%) angiosarcomas, 1 of 61 (1.6%) soft tissue leiomyosarcomas (this was the same case that was positive for DOG1.1), 9 of 11 (81.8%) breast fibroadenomas (expression restricted to myoepithelial cells), and 3 of 4 (75.0%) adenoid cystic carcinomas of the head and neck (Tables 3, 4).

DISCUSSION The diagnosis of most GISTs can be made based on the combination of clinical, histologic, and immunohistochemical features, mainly the expression of KIT/ CD117.15 Up to 5% of GISTs, however, are negative for KIT expression.9,10KIT and PDGFRA mutation status can help in establishing the diagnosis of GIST in this setting, but mutation screening is an expensive and timeconsuming approach that may be better reserved for directing tyrosine kinase inhibitor therapy.16 DOG1 is a protein encoded by TMEM16A (also known under several names that include TMEM16, FLJ10261, ANO1, ORAOV2, and TAOS2), a gene located within the CCND1-EMS1 locus on human chromosome 11q13 that is amplified in esophageal, bladder, and breast cancers.17 On the basis of a geneexpression profile using cDNA microarrays, TMEM16A was found to be up-regulated in both KIT and PDGFRAmutant GISTs.11 In separate studies, researchers found

TABLE 2. DOG1.1 and K9 Positivity in GISTs Including KIT-negative GISTs

All GISTs (n = 668) KIT-negative GISTs (n = 25)

DOG1.1-positive GISTs Positive Cases (%)

K9-positive GISTs Positive Cases (%)

538 (80.5) 5 (20.0)

642 (96.1) 19 (76.0)

DOG1 indicates discovered on GIST-1; GIST, gastrointestinal stromal tumor.

r

2010 Lippincott Williams & Wilkins

that TMEM16A encodes a calcium-regulated chloride channel protein.18,19 West et al11 reported that DOG1 was superior in sensitivity and specificity to KIT/CD117 using a rabbit polyclonal antibody (S284), being expressed in 97.8% of GISTs. Later, the same group presented similar results with 2 new mouse monoclonal antibodies against DOG1 (DOG1.1 and DOG1.3), with DOG1.1 showing advantages over DOG1.3. DOG1.1 was positive in 87% of GISTs, whereas KIT/CD117 was expressed in 74% of tumors. Moreover, DOG1.1 rarely was expressed in tumors other than GISTs, including 1of 326 leiomyosarcomas, 1 of 39 synovial sarcomas, 1 of 10 desmoplastic melanomas, and some carcinomas. In contrast, KIT/ CD117 was expressed in several different tumors, including vascular neoplasms, extraskeletal myxoid chondrosarcoma, Ewing sarcoma/primitive neuroectodermal tumor, clear cell sarcoma, desmoplastic small round cell tumor, alveolar soft part sarcoma, carcinomas from different sites, melanoma, and germ cell tumors.12 Liegl et al13 demonstrated that DOG1.1 was a sensitive immunohistochemical marker for GIST, comparable with KIT, with the additional benefit of detecting 36% of KIT-negative GISTs. Miettinen et al14 recently first used the K9 antibody to investigate DOG1 expression

TABLE 3. Nonmesenchymal Tumors Included in this Study and Results for DOG1.1 and K9 Immunostaining Tumors Melanoma Fibroadenoma of breast Non-Hodgkin lymphoma Small cell carcinoma Carcinoma of lung Carcinoma of kidney Carcinoma of thyroid (papillary) Thymoma Hodgkin lymphoma Carcinoid Adenoid cystic carcinoma of head and neck Adamantinoma Phyllodes tumor Carcinoma of breast

No. Cases

DOG1.1

K9

15 11 7 6 5 5 5 5 5 4 4

0 0 0 0 0 0 0 0 0 0 0

0 9 (81.8%)* 0 0 0 0 0 0 0 0 3 (75.0%)

2 2 2

0 0 0

0 0 0

*Expression restricted to myoepithelial cells.

www.appliedimmunohist.com |

335

Appl Immunohistochem Mol Morphol

Lopes et al

TABLE 4. Mesenchymal Tumors Other Than Gastrointestinal Stromal Tumor Included in this Study and Results for DOG1.1 and K9 Immunostaining Tumors Soft tissue undifferentiated pleomorphic sarcoma Soft tissue leiomyosarcoma Synovial sarcoma Atypical lipomatous tumor and lyposarcoma Soft tissue fibromatosis (desmoid-type fibromatosis) Soft tissue schwannoma Soft tissue and bone tenosynovial giant cell tumor Solitary fibrous tumor Rhabdomyosarcoma Uterine leiomyoma Ewing sarcoma/primitive neuroectodermal tumor Bone osteosarcoma Angiosarcoma Endometrial stromal sarcoma Soft tissue epithelioid hemangioendothelioma Neurofibroma Soft tissue extraskeletal myxoid chondrosarcoma Soft tissue nodular/proliferative fasciitis Ovary fibroma Dermatofibrosarcoma protuberans Gastrointestinal leiomyosarcoma Fibrous dysplasia Malignant peripheral nerve sheath tumor Soft tissue myxofibrosarcoma Gastrointestinal leiomyoma Angiomyolipoma Bone pigmented villonodular synovitis Soft tissue myxoma Bone giant cell tumor Soft tissue fibroma of tendon sheath Soft tissue desmoplastic small round cell tumor Soft tissue glomus tumor Soft tissue and bone non-ossifying fibroma Digital fibromatosis Genital tract leiomyosarcoma Soft tissue inflammatory myofibroblastic tumor Atypical fibroxanthoma Granular cell tumor Soft tissue chondrosarcoma Bone enchondroma Gastrointestinal schwannoma Soft tissue leiomyoma Neuroblastoma Soft tissue hemangioma Soft tissue lymphangioma Soft tissue epithelioid sarcoma Soft tissue low-grade fibromyxoid sarcoma Kaposi sarcoma Soft tissue alveolar sarcoma Soft tissue chondromyxoid fibroma Soft tissue desmoplastic fibroma Soft tissue clear cell sarcoma Soft tissue hibernoma Aneurysmal bone cyst Bone fibrous histiocytoma Ganglioneuroma Soft tissue ossifying fibroma Soft tissue deep aggressive angiomyxoma Juvenile xanthogranuloma Soft tissue calcifying aponeurotic fibroma Soft tissue collagenous fibroma Elastofibroma

336 | www.appliedimmunohist.com

No. Cases 72 61 57 43 33 30 26 23 20 19 17 15 14 13 12 12 11 11 11 11 9 9 9 8 8 8 8 7 7 7 7 6 6 6 5 5 5 4 4 4 4 4 3 3 3 3 3 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1

DOG1.1

K9

0 0 1 (1.6%) 1 (1.6%) 2 (3.5%) 5 (8.8%) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 (7.1%) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0



Volume 18, Number 4, July 2010

in GISTs, tumors other than GISTs, and normal tissues. They found that the overall sensitivity of K9 and KIT in GISTs was nearly identical (94.4% and 94.7%) and that only 2.6% of GISTs were negative for both KIT and K9, results similar to the 0.9% of cases presented in this study. The expression of DOG1 using K9 was found by those authors to be highly specific for GIST, even though rare smooth muscle tumors and synovial sarcomas also were positive. DOG1 expression also was observed in some carcinomas of the gastrointestinal tract. Interestingly, those authors found K9 to be positive on the luminal surface of the gastric mucosa, salivary gland, pancreatic acini, and intrahepatic bile ducts; gallbladder glandular elements and breast and prostatic myoepithelial/basal cells were also variably positive. In our study, K9 positivity was observed in gallbladder and in myoepithelial cells of breasts with fibroepithelial lesions. In this study, we demonstrated that the new commercially available antibody against DOG1, K9, is as sensitive as KIT/CD117 (96.1% using K9 and 96.3% using KIT) for the diagnosis of GIST, whereas the antibody DOG1.1 was positive in 80.5% of tumors. Moreover, K9 was positive in 76% of KIT-negative GISTs, whereas DOG1.1 was positive in only 20% of KIT-negative tumors. Only 0.9% of GISTs included in this study were negative for KIT, DOG1.1, and K9. Even though K9 was more sensitive than DOG1.1 for the diagnosis of GIST, the commercially available antibody was less specific: K9 was positive in 5 of 57 synovial sarcomas, 1 of 14 angiosarcomas, 1 of 61 soft tissue leiomyosarcomas, and 3 of 4 adenoid cystic carcinomas of the head and neck, whereas DOG1.1 was positive in 2 of 57 synovial sarcomas and 1 of 61 soft tissue leiomyosarcomas only. In conclusion, the commercially available K9 antibody against DOG1 protein has sensitivity similar to KIT for the diagnosis of GIST, being of optimal utility for the diagnosis of most KIT-negative GISTs. Moreover, the combination of both KIT and K9 antibody in a panel of immunohistochemistry can define the diagnosis of GIST in more than 99% of cases. As a result of such findings, KIT and PDGFRA mutation screening for diagnostic purposes should be reserved for those rare KIT and K9-negative cases and whenever tyrosine kinase inhibitor therapy is considered for the management of the disease. Caution with the use of K9 antibody should be considered because of its possible positivity in neoplasms other than GISTs, mainly synovial sarcomas, angiosarcomas, and leiomyosarcomas, which enter in the differential diagnosis of GIST. REFERENCES 1. Rubin BP. Gastrointestinal stromal tumours: an update. Histopathology. 2006;48:83–96. 2. Miettinen M, Lasota J. Gastrointestinal stromal tumours. Review on morphology, molecular pathology, prognosis, and differential diagnosis. Arch Pathol Lab Med. 2006;130:1466–1478. 3. Hornick JL, Fletcher CD. The role of KIT in the management of patients with gastrointestinal stromal tumors. Hum Pathol. 2007;38: 679–687. r

2010 Lippincott Williams & Wilkins

Appl Immunohistochem Mol Morphol



Volume 18, Number 4, July 2010

4. 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. 5. Corless CL, McGreevey L, Haley A, et al. KIT mutations are common in incidental gastrointestinal stromal tumors one centimeter or less in size. Am J Pathol. 2002;160:1567–1572. 6. Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science. 2003;299: 708–710. 7. Demetri GD, Von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med. 2002;347:472–480. 8. Demetri GD, Van Oosterom AT, Garrett CR, et al. Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomized controlled trial. Lancet. 2006;368:1329–1338. 9. Medeiros F, Corless CL, Duensing A, et al. KIT-negative gastrointestinal stromal tumors: proof of concept and therapeutic implications. Am J Surg Pathol. 2004;28:889–894. 10. Lopes LF, Ojopi EB, Bacchi CE. Gastrointestinal stromal tumor in Brazil: clinicopathology, immunohistochemistry, and molecular genetics of 513 cases. Pathol Int. 2008;58:344–352. 11. West RB, Corless CL, Chen X, et al. The novel marker, DOG1, is expressed ubiquitously in gastrointestinal stromal tumors irrespective of KIT or PDGFRA mutation status. Am J Pathol. 2004;165:107–113.

r

2010 Lippincott Williams & Wilkins

Comparison of DOG1 Antibodies

12. Espinosa I, Lee CH, Kim MK, et al. A novel monoclonal antibody against DOG1 is a sensitive and specific marker for gastrointestinal stromal tumors. Am J Surg Pathol. 2008;32:210–218. 13. Liegl B, Hornick JL, Corless CL, et al. Monoclonal antibody DOG1.1 shows higher sensitivity than KIT in the diagnosis of gastrointestinal stromal tumors, including unusual subtypes. Am J Surg Pathol. 2009;33:437–446. 14. Miettinen M, Wang ZF, Lasota J. DOG1 antibody in the differential diagnosis of gastrointestinal stromal tumors: a study of 1840 cases. Am J Surg Pathol. 2009;33:1401–1408. 15. Fletcher CD, Berman JJ, Corless C, et al. Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol. 2002;33: 459–465. 16. Blay JY. New paradigms in gastrointestinal stromal tumour management. Ann Oncol. 2009;20:i18–i24. 17. Katoh M, Katoh M. FLJ10261 gene, located within the CCND1EMS1 locus on human chromosome 11q13, encodes the eighttransmembrane protein homologous to C12orf3, C11orf25 and FLJ34272 gene products. Int J Oncol. 2003;22:1375–1381. 18. Caputo A, Caci E, Ferrera L, et al. TMEM16A, a membrane protein associated with calcium-dependent chloride channel activity. Science. 2008;322:590–594. 19. Yang YD, Cho H, Koo JY, et al. TMEM16A confers receptoractivated calcium-dependent chloride conductance. Nature. 2008; 455:1210–1215.

www.appliedimmunohist.com |

337