Analysis of the amount of tissue sample necessary for mitotic count and Ki-67 index in gastrointestinal stromal tumor sampling.

There are no established opinions concerning whether the amount of tissue affects the accuracy of histological analyses in gastrointestinal stromal tumors (GISTs). The aim of the present study was to investigate the appropriate amount of tissue sample needed for mitotic count based on the risk classification of GISTs and the Ki-67 index using the following three methods: endoscopic ultrasound-guided fine-needle aspiration (FNA), a novel sampling method called tunneling bloc biopsy (TBB), and biopsy forceps followed by TBB (Bf). Forty-three samples (12 FNA, 17 TBB and 14 Bf) diagnosed as GISTs by immunohistological analysis were utilized. The major and minor axes and overlay area of one piece of specimen (OPS) from the three sampling methods were measured using digital imaging software and were analyzed comparatively regarding the acquisition of histological data. The mean major and minor axes (mm) and overlay areas (mm2) were in the order of TBB > Bf > FNA. The evaluable rates by mitotic count and Ki-67 were, respectively, 75% (9/12) and 83.3% (10/12) for FNA samples, 100% (17/17) and 100% (17/17) for TBB samples, and 100% (14/14) and 100% (14/14) for Bf samples (P>0.05). Three FNA samples were judged unevaluable due to too small specimens in overall diagnosis including mitotic count and Ki-67, calculating the cut-off value for the overlay area of OPS as 0.17 mm2. Comparing the concordance rates between the pre- and post-operative samples, TBB samples was significantly better than FNA (P<0.05). Conclusively, while the amounts of tissues obtained by TBB and Bf are unnecessary for the histological assessment of mitotic count and Ki-67 index, developments of the FNA method are needed to minimize sample error. Considering the technical aspects, as well as the size of the specimens, could help to guide therapeutic planning and improve diagnostic yield for GI subepithelial tumors.