A48
GIST mutation analysis in from EUS FNA cytology material: correlation with resected tumours in 10 cases
Brendan O'Sullivan1, Nayneeta Deshmukh1, Jenny Bell2, Rahul Hejmadi1, William Cuthbertson1, Brinder Mahon1, Philippe Taniere1
1University Hospital Birmingham NHS Foundation Trust, Birmingham, UK, 2Birmingham Women's Hospital Foundation Trust, Birmingham, UK
Background
Endoscopic ultrasound scan (EUS) is an efficient technique for obtaining cytology samples from soft tissue tumours situated along the upper gastrointestinal tract. C-KIT and PDGFR alpha gene mutation analysis represents a helpful supplementary test in diagnosing gastrointestinal stromal tumours (GIST) and also in predicting response to targeted drugs such as Imatinib. We assessed the reliability of mutation analysis in cytology specimens by comparing results with those from tumour resection specimens.
Method
We proceeded with the analysis of 14 cases from surgical EUS fine needle aspiration (FNA) cytology material. DNA was extracted from the needle washings preserved in fixative (CRR); direct DNA sequencing was subsequently performed for exons 9, 11, 13 and 17 of the C-KIT gene and exons 12, 14 and 18 of the PDGFRA gene. Eight/fourteen patients had surgical resection to date. The remaining 6 cases were therefore excluded from our final analysis.
Results
We obtained good quality DNA and PCR products in 13/14 cases. Among the 8 paired cases, 5 showed absolute concordance in terms of mutation status (all these cases showed a mutation in exon 11 of C-KIT gene: there were three missense, one insertion and one deletion). However 3 cases showed a mutation in the resected tumour not seen in the previously tested cytology specimen.
Conclusion
Our study shows that EUS FNA material is of good quality for retrieving DNA and represents a valuable material for molecular testing of tumours with high specificity. This can therefore be used in routine practice to avoid more aggressive investigation if mutation analysis is required. However, caution should be taken in interpreting a wild-type result which is potentially to be due to poor sampling.
