The Statistical Analysis of the EUS-Features of Insulinomas

At the beginning of our research we were guided by the accepted at that time description of insulinoma as roundish hypoechoic lesion. While accumulating personal experience we have found out the big variety of the EUS images of insulinomas.

Aim: to define frequency of occurrence of the basic EUS features of insulinomas.

Materials and methods: Preoperative endoscopic ultrasound (EUS) has been done during 7 years (2003–09) for 61 patients (12 men, 49 women; mean age 43 years, range 12 to 69 years) with sporadic insulinomas (46) and associated with multiple endocrine neoplasia (MEN) syndrome (15). Among them there were 6 patients, ineffectively previously operated in other centers. All patients were operated with the consecutive immunohistochemical studies for surgical specimens - 66 insulinomas were detected (uncinate process 9, head 20, neck 6, body 13, tail 18). The mean size of these tumors was 17mm (range 9 to 35mm). Preoperative EUS has revealed 62 insulinomas, videorecordings of failed EUS cases have been kept. The ultrasound portrait of each insulinoma has been created after the analysis of a contour, homogeneity and level of echogenicity, inclusions and the form of insulinomas. The basic echo-types of insulinomas have been allocated under ultrasonic characteristics. I Echo-type: hypoechoic homogeneous-, II Echo-type: isoechoic homogeneous-, III Echo-type: heteroechogeneous formation of circular, oval or lobular shape, with clear contour; IV Echo-type includes focuses without clear contour. Also the size and localization of insulinomas were considered.

Results: The percentage parity of I-II-III-IV Echo-types in the general group of 66 insulinomas was: 39% (26)-23% (15)-28% (18)-11% (7), among previously operated patients: 33% (2)-33% (2)-17% (1)-17% (1), among 4 insulinomas not revealed during EUS (on the saved videorecordings): 0% (0)-25% (1)-25% (1)-50% (2). The percentage parity of I-II-III-IV Echo types depending on the size of a tumor was: less than 1cm (6) 33.3% (2) -0% (0)- 33.3% (2) - 33.3% (2); 1–1.5cm (17) 41% (7) - 18% (3) - 29% (5) - 12% (2); 1.5–2cm (31) 42% (13) - 32% (10) - 16% (5) - 10% (3); 2–2.5cm (9) 44.3% (4) - 22.3% (2) - 33.3% (3) - 0% (0); more than 2.5cm (3) 0% (0) -0% (0) -100% (3) -0% (0).

Conclusion: Despite prevalence of the I (hypoechoic) Echo-type of insulinomas, it is necessary to consider other options to increase their detectability. Considering small frequency of occurrence of the given pathology, it is expedient to carry out the large-scale international analysis of the kept video data of EUS of insulinomas.