| Abstract: |
The nonepithelial, nonlymphoid tumors of the gastrointestinal tract are heterogeneous in terms of clinical presentation, behavior, pathology, and genetic features. Concepts regarding these tumors have changed rapidly over the past decade as nomenclature has evolved. Many of these tumors have no muscle differentiation, and designations such as leiomyoma or leiomyosarcoma are inappropriate for many of these neoplasms. With an improved understanding of the biology of these tumors, gastrointestinal stromal tumor (GIST) is used as a specific term for tumors of the gastrointestinal tract that lack markers of myogenic differentiation, but stain positive for vimentin, and express CD34 and CD117, the product of the c-kit oncogene. Both benign and malignant types are recognized. In addition to myogenic tumors and GIST, gastrointestinal autonomic nerve tumors (GANT) are also recognized. Complete en bloc surgical resection, when possible, is the cornerstone of therapy. Metastasis tends to occur to the liver and within the peritoneal cavity, especially in patients whose tumors have ruptured spontaneously or been violated by the surgeon. Incomplete surgical resection and metastatic disease indicate a dismal prognosis in the majority of patients. Recurrent or metastatic disease is often resected, but this has an uncertain impact on outcome. Operation may palliate patients with intestinal obstruction or other symptoms. For patients with unresectable disease, the results with systemic chemotherapy have been dismal. Treatment with doxorubicin/ifosfamide combinations is of dubious value. Hepatic arterial embolization, with and without intra-arterial chemotherapy, results in regression of liver metastases in selected patients. Regression has also been seen using intrahepatic arterial infusion of doxorubicin without embolization. The impact of such treatment on outcome, however, is poorly studied. Aggressive surgical resection of peritoneal metastases with intraperitoneal chemotherapy has been advocated, but requires formal study in large trials. |
