1. ¹Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, Washington, DC, USA
2. ²Department of Biology and Genetics, Medical University of Gdansk, Gdansk, Poland
3. ³Department of Pathology, Holycross Cancer Center, Kielce, Poland
4. ⁴Department of Pathology, Saitama Medical School, Saitama, Japan
5. ⁵Department of Pathology, Haartman Institute of the University of Helsinki, Helsinki, Finland
6. ⁶Veteran Affairs Medical Center and Institute of Molecular Medicine and Genetics and Department of Pathology, Medical College of Georgia, Augusta, GA, USA
7. ⁷Department of Pathology, Otto-von-Guericke University, Magdeburg, Germany
8. ⁸Department of Pathomorhpology, Jagiellonian University, Krakow, Poland

Correspondence: Dr J Lasota, MD, Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, 14th Street and Alaska Avenue, N.W., Washington, DC 20306-6000, USA. E-mail: lasota@afip.osd.mil

^*Research fellows at the Department of Soft Tissue Pathology, AFIP, Washington, DC, USA.

Received 8 September 2004; Revised 26 October 2004; Accepted 28 October 2004; Published online 6 December 2004.

Abstract

Mutational activation of KIT or PDGFRA is considered an early step in pathogenesis of gastrointestinal stromal tumors (GISTs); however, other nonrandom genetic changes have also been identified. At least three common regions of deletions on chromosome 22q, which may harbor putative tumor suppressor genes, have been defined. However, mapping of these regions has been inconsistent. It has also been speculated that GI autonomous nerve tumors (GANTs), GISTs with ultrastructural features suggestive of autonomic nerve differentiation, are characterized by a specific deletion involving 22q13 cytogenetic region. This study was undertaken to evaluate loss of heterozygosity (LOH) on chromosome 22q in 50 GISTs, including 10 GANTs. Four tumors were incidental minimal lesions 10 mm in diameter. LOH was evaluated using 20 PCR-based microsatellite markers and capillary gel electrophoresis. In all, 15 (30%) cases showed LOH of more than 75% of informative markers, suggesting loss of chromosome 22q. A total of 24 GISTs (50%) revealed LOH of one to seven informative markers clustered in different loci suggesting simultaneous involvement of different regions. The highest frequency of LOH was seen at D22S922 and D22S425, mapped to 22q13.33 and 22q11.22, respectively. However, LOH at other regions including IL2RB and NF2 locus was also found. No NF2 mutations were identified in four analyzed tumors. LOH on chromosome 22q was more frequent among intestinal than among gastric GISTs; however, there was no difference between LOH pattern seen in tumors defined by different histologic, ultrastructural (GANT) and molecular features (KIT and PDGFRA mutations). Although minimal GISTs revealed LOH on chromosome 22q, there was a higher LOH frequency in malignant than in benign tumors. An isolated LOH at D22S425 was equally found in both benign and malignant tumors. These observations may suggest that LOHs on chromosome 22q in GISTs play a role in early stages of tumor formation as well as in late tumor progression.