A prospective, controlled and blinded trial of 43 patients a with a histologic diagnosis of gastric cancer, data was collected consecutively.

The aim was a pretherapeutic evaluation of TNM-stage and an resectability assessment by laparoscopic ultrasound (LUS). Non resectability was defined as peritoneal carcinosis, ingrowth in the pancreas, aorta, celiac trunc, liver and metastasis.

Patients and Methods. Twenty-three(54%) patients had a tumor located in the upper part of the gastric body, seven patients(16%) in the mid- or lower part of the gastric body. Nine patients(21%) was in the antrum and in 2 (5%) the whole stomach was infiltrated. All patients had adenocarcinomas, but 11(26%) had low differentiated type. Exclusion criteria was poor general condition, suspicion of severe adhesions, known metastasis, competitive disease that would contra-indicate laparotomy and lack of informed consent. All patients had a chest x-ray, external ultrasound scanning, CT-scanning and endoscopic ultrasound examination performed.

Results. The overall accuracy of T-staging was 0.81 (0.66-0.91) confidence interval n=34/42. Four patients(10%) was overstaged, and two patients(5%) was understaged. Three patients(7%) was not assessed by LUS, because of small size and localization. The N-stage accuracy was 0.67 (0.48-0.82) n=22/33. The M-stage was evaluated in 41 of the 43 patients (95%). The M-stage accuracy was 0.88 (0.74-0.96) n=36/41. The overall TNM-stage accuracy was 0.59 (0.42-=.74) The resectability assessment in 40/43 patients with an accuracy of 0.80 (0.64-0.91) n=32/40. The sensitivity and specificity of non-resectability 1.00 and 0.96.

Discussion. The failures of laparoscopic ultrasound staging was primarily due to wrong T- and M-stage. Laparoscopy was able to correct staging in the two patients with wrong M-stage. The accuracy of TNM-stage may rise by adding biopsy. The data confirm that laparoscopy and LUS can stage T3 and T4 tumors accurately. The pretherapeutic assessment of resectability may be beneficial in the therapeutic decision planning. The results indicate that by LUS one may avoid laparotomy in 37% and in other studies by 19-44%.