Laboratory Investigation
United States and Canadian Academy of Pathology The United States and Canadian Academy of Pathology
LWW Lippincott Williams and Wilkins
publishes Laboratory Investigation
--
                                      View Future Titles
Through Mar 2001
       Archives
Aug 1965 - Feb 2001
       Search Articles
Aug 1965 - Feb 2001
       Browse by Subject
Aug 1965 - Feb 2001
                      
Instructions to authors

Subscriptions

About the journal
   
  Loss of Heterozygosity at 1p36 Predicts Poor Prognosis in Gastrointestinal Stromal/Smooth Muscle Tumors
Editorial board

Email alerts

'Net Tips

Help

Feedback

Guestbook








  Timothy O'Leary, Susanne Ernst, Ronald Przygodzki, Theresa Emory, and Leslie Sobin
   
  Departments of Cellular Pathology (TO, SE, RP) and of Hepatic and Gastrointestinal Pathology (TE, LS), Armed Forces Institute of Pathology, Washington, DC
   
  SUMMARY: Gastrointestinal stromal/smooth muscle tumors are uncommon neoplasms for which current criteria for diagnosing malignancy (size and mitotic index) sometimes fail to predict outcome. Cytogenetic studies reveal frequent chromosome 1 abnormalities in these tumors, but significant underlying molecular changes have not been elucidated, and their significance is unknown. DNA was obtained from the formalin-fixed, paraffin-embedded tissue of 80 gastrointestinal stromal/smooth muscle tumors. Tumors were topographically microdissected from surrounding normal tissue; microsatellite markers from tumor and normal tissue were amplified by PCR in the regions of chromosome 1p36 (D1S199, D1S228, D1S450, D1S214, D1S243), 1p12 (D1S418), 1p13 (D1S252, D1S514), and 1q32 (D1S103). The presence or absence of heterozygosity for each case was mapped at each informative marker. Relationships among loss of heterozygosity (LOH), tumor size, mitotic index, and survival were investigated using correlation analysis, Kaplan-Meier plots, and the Cox model. LOH at 1p36 was found in 24 of 80 cases, suggesting the possibility of a tumor suppressor gene at 1p36 near the site of a suspected neuroblastoma tumor suppressor gene. Patients whose tumors demonstrated LOH at 1p36 had significantly shorter survival (p = 0.017) than those whose tumors did not. LOH at 1p36 retained independent prognostic significance in a multivariate model that included KIT mutation status and tumor size; the mitotic index, however, did not retain independent significance in such a model. LOH was observed at 1p12-1p13 (most frequently at 1p13.3) in 19 of 80 cases, but loss of heterozygosity at this site did not influence survival. No LOH was observed near 1q32. These findings provide strong evidence for a prognostically significant tumor suppressor gene in the region of chromosome 1p36.3.