Mutation analysis of gastrointestinal stromal tumors – importance and preliminary results of the National Center for GISTs diagnosis in Slovakia

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract characterized by expression of CD117 and harboring activating mutations in KIT and PDGFRA proto-oncogenes. It is a good example of solid tumor to apply our understanding from aberrant signal transduction to drug development. The constitutive activated KIT and PDGFRA oncoproteins serve as crucial diagnostic and therapeutic targets. The aim of this presentation is to show the importance of the mutational analysis in GISTs, the steps leading to the establishment of the algorithm of molecular examinations and present the preliminary results from molecular analysis of primary mutations in GISTs.

The receptor tyrosine kinases (RTKs) KIT and PDGFRA belong to the type III receptor family. The common topology of RTKs includes an extracellular domain, a single transmembrane segment and a cytoplasmic, juxtamembrane and kinase domains. Gain-of-function activating mutations in KIT occur in most GISTs, and a majority of them have been found within the juxtamembrane domain in coding exon 11. In-frame deletions and base substitutions are the most common mutations in exon 11 of KIT. In exon 9 of KIT, a very common mutation is represented by a six nucleotides insertion that results in duplication of amino acid alanine and tyrosine. The other primary KIT mutations have been identified within exons 13 and 17. In the case of KITnegative GIST, activating mutationshave been found in the PDGFRA gene within exons 12, 14 and 18. There are controversies regarding the prognostic value of KIT exon 11 mutations in GISTs. In early studies, it was shown that KIT exon 11 mutations are more common in large mitotically active tumors. Other studies detected KIT mutations as ubiquitous feature of GISTs found in malignant and benign tumors too. The presence of the most common deletion p.W557_K558del has been associated with negative prognosis. The position of the primary mutation can be predictive for response to the imatinibmesylate. Clinical studies have reported that patients with KIT exon 11 showed better response to the imatinibmesylate treatment than patients with KIT exon 9, 13 and 17 mutations and wildtyp. Patients with exon 9 mutations require a higher dosage to achieve similar response as patients with exon 11 mutations. During the treatment, the patients can develop secondary resistance based on the secondary mutations within KIT and PDFGRA, KIT amplification or activation of other TKs.

To improve identification of the diseaseand understanding of patients survival,a GIST clinical registry was founded ((http://gist.registry.cz/) for patients from Czech Republic and Slovakia where clinical, histologicaldata and results from gene mutation analyses are collected. In Slovakia, the center for identification and assessment of GIST based on immunohistochemical reaction with CD117, morphologicaland molecular characteristics was established at the Department of Pathological Anatomy andthe Department Molecular Biology in Martin in collaboration with Genetoncompany.

A diagnosis of GIST requires a multidisciplinary approach of clinical, pathological and genetic examinations. Consensus guidelines for GIST prognosis defined at the National Institutes of health GIST Workshop in 2001 emphasize tumor size, mitotic index and GIST localization were applied for risk stratification of primary tumors. For molecular analyses, the DNA was extracted from FFPE sections after deparaffinization in xylene and rehydratation through a series of descending concentrations of alcohol. The primers were generated for exons 9, 11, 13 and 17 of KIT and for exons 12, 14 and 18 of PDGFRA based on the reference sequences from the GeneBank NG_007456.1 and NG_009250.1, respectively. SSCP analysis on the nondenaturing PAGE was used for mutational screening in first analyses. The purified PCR products from the PCR analysis were directly sequenced using the Big Dye Terminator kit v.3.1. Nomenclature of the identified mutations is based on the HUGO recommendations based on human KIT (X06182) and PDFGRA (M12574) mRNA sequences from the GeneBank and the sequences were evaluated using Chromas and SeqScape software.

The SSCP analysis was time consuming and inefficient for the mutational screening and the direct DNA sequencing was the first method of choice for the molecular analyses. Based on the mutation frequencies the mutational analyses were started with exons 11 and 9 of KIT and 18 of PDFGRA. In case of negative results, the sequencing of other exons followed.

From the 235 examined tissues from patients with GIST, in 192 cases a primary mutation in one of the tested exons was identified (81, 7%). Oftotal 192 mutations, 160 (83%) were found in KIT and 32 (17%) in PDGFRA. The most common mutations in KIT were localized within exons 11 and 9. There were identified two mutations in exon 13 and one in 17.The most common mutations in exon 11 were amino acid deletions (61%). After dividing the deletions and delins mutations into two groups - the small (up to 20 bp) and large (more than 20bp) deletions it was obvious that small deletions occur mostly at 5’ end of exon 11 between codons 550 and 562. Deletions larger than 20 bp are towards to the 3’end ofexon 11 betweencodons 557 to 580. The most frequently deleted was codon 570. The most prevalent small deletion was p.W557_K558del (20%).In the PDGFRA gene, the most frequent were mutations in exon 18 (72%).

In our study, we identified mutations in 81,7%. The similar studies from Norway and Iceland reported previously 75, 7% and 92, 9% frequencies respectively. The differences can be results of different cohort constitution concerning the malignity of tumors. The other aspect is the quality of the FFPE tissues where it was shown that the mutation detection in older tissues is less efficient than when working with more recent blocks. The screening method which is more sensitive than direct DNA sequencing can have effect on the mutation detection too. The proportion of different mutations in exons 11 and 9 of KIT and exon 18 of PFGFRA are comparable to previously published reports. Based on mutation analysis, we identified patients who benefit from higher dosage of imatinibmesylate. Our future efforts will focus on detection secondary mutations after treatment.