Comparison of EUS-Guided Trucut with EUS-Guided Fine-Needle Aspiration in Subepithelial Tumors: Preliminary Results of a Prospective Study

Background: Subepithelial tumors are difficult to diagnose on the basis of cytology alone. Aim: to prospectively compare diagnostic accuracy of EUS-guided trucut needly biopsy (EUS-TNB) with EUS-guided fine-needle aspiration (EUS-FNA) in the diagnosis of subepithelial tumors. Patients and Methods: we included ten consecutive patients who underwent both EUS-TNB and EUS-FNA of a subepithelial lesion. Inclusion criteria were: 1. diameter >2cm; 2. solid lesion and 3. Platelet count >50,000 and protrombine time >50%. The order of EUS-TNB and EUS-FNA was randomly assigned. All procedures were performed with the patient under conscious sedation. The biopsy was obtained with a disposable 19-gauge trucut biopsy needle and EUS-FNA with a disposable 22-gauge needle. All specimens were evaluated by one on-site cytopathologist who assessed the quality of the specimen. The patient underwent as many passes as necessary until a complete cylinder was identified or until the cytopathologist informed that an adequate specimen was obtained for EUS-TNB and EUS-FNA respectively, with a maximum of three passes in both cases. Final diagnosis was based on cytology or histology and immunohistochemical determinations when appropriated. Results: Ten tumors were sampled with a mean size of 58 (+61) x 37 (+ 27)mm. All lesions were in the stomach. The mean number of passes was 1.5+0.7 with EUS-TNB and 1.7+0.8 with EUS-FNA (p=ns). Final diagnosis was stromal tumor (n=8) and carcinoma (n=2). EUS-TNB provided with the diagnosis in 7/10 cases (70%) and immunohistochemical determination of C-kit was possible in 6/6 tumors in which it was indicated (100%). EUS-FNA provided with the diagnosis in 9/10 cases (90%, p=ns), however, the immunohistochemical determination of C-kit was only possible in 5/9 tumors in which it was indicated (55%, p=0.056). Device failure occurred in two out of the three tumors not diagnosed by trucut. EUS-FNA was possible in all of them. There were no complications. Conclusions: Determination of c-kit seems to be more reliable in the tissue specimen provided by EUS-TNB. EUS-FNA can serve as a rescue technique in cases of Trucut failure. More patients have to be studied before taking definitive conclusions about final accuracy of the two techniques.