Dog 1 Expression In Gastro-intestinal Stromal Tumors Reported In Nairobi An Immuno-histochemical Evaluation of a Novel Antibody in Comparison to Morphology & CD117 Staining
Background: All Gastro-Intestinal Stromal Tumors (GISTs) should be diagnosed as they are amenable to treatment with Tyrosine Kinase Inhibitors (TKI). TKIs are a group of targeted therapies, which have revolutionalised treatment of tumors such as Chronic Myeloid Leukemia (CML) and more importantly GISTs. A commonly used TKI is Imatinib whose common brand name is Gleevec or Glivec. GISTs present diagnostic challenges on routine Hematoxylin and Eosin (H&E) stain and on Immuno-histochemistry with the commonly used markers in Kenya; CD117 and CD34. Due to these challenges a lot of research efforts are directed at markers that will identify all GISTS. Several markers have been shown to improve diagnostic sensitivity and specificity when used alone or together with CD117. Discovered on GIST 1(DOG 1) antibody is one such marker. Antibodies to DOG 1 protein have been shown to improve diagnostic sensitivity and specificity when included in the Immuno-Histo Chemical (IHC) panel for GIST. Objective: To evaluate DOG 1 expression in tumors diagnosed as GISTs on routine histology and CD117 staining and in those suspected to be GISTs on morphology alone. Design: Cross-sectional descriptive study Setting: Kenyatta National Hospital, Nairobi Hospital and Pathologists, Lancet Kenya Laboratory. Main outcome measures: Fifty three previously diagnosed GISTs were analyzed for DOG 1 expression by immunohistochemistry on paraffin wax embedded tumor tissue blocks. DOG 1 expression was correlated with age, sex, and anatomic site of tumor, mitotic counts and CD117 expression. Results: The males were 25(47.2%) while the females were 28(52.8%).M: F was 9:10.Mean age was 50.5 years while the median age was 53 years. The odds of a tumor being DOG 1 positive test in a male patient were 3.5 higher than a negative test .The odds of a tumor being DOG 1 positive are lower for tumors with more than 10 mitoses in 50 hpf than negative (OR =0.2). Tumors in males had more odds of being positive for CD 117(OR 4.6) when compared to those in females. Mitoses and tumor size had little correlation with CD117 positivity or negativity. The cases were stratified using DOG 1 and CD117 immunohistochemistry into GISTs and non- GISTs. The M: F in GIST cases is 1.3:1, median age was 52.3 years. Thirty three (62.3%) tumors were CD117 positive while 35(66%) tumors were positive for DOG 1 staining. Thirty one (58.5%) tumors were positive for both markers, 16 (30.2%) were negative for both markers. Six (11.3%) tumors were positive for only one marker; two (3.8%) with CD117 and 4(7.5%) with DOG 1. Using both CD117 and DOG1, 37(69.8%) 16(30.2%) found to be negative for both markers were classified as non-GIST. Conclusion: GIST in Kenya occurs at an earlier age with an almost equal sex distribution. DOG 1 immuno-positivity was positively correlated with the male sex as was CD117 expression. Although DOG 1 identified more cases than CD 117 the difference was not statistically significant. A significant proportion of tumors in this study were negative for both DOG 1 and CD117; these are thought to represent other biological entities. Extra gastro-intestinal GISTs were common in this study which may be attributed to late patient presentation when the tumor has already metastasized. All the EGISTs in this study were intra-abdominal; no extra-abdominal cases were described. Recommendations: Epidemiological long term prospective studies should be done to ascertain the clinical behavior of GISTs including age of onset in Kenya. DOG 1 and CD117 should be used together to diagnose more GIST cases. An immuno-histochemical study using a broad panel of antibodies should be done on the non- GIST cases in this study to determine their true diagnoses. The apparent high frequency of metastatic GISTs needs to be addressed by advising clinicians to have a high index of suspicion and diagnose GISTs early when they are more amenable to cure.