TNM staging and assessment of resectability in esophageal,gastric and pancreatic cancer using Laparoscopic Ultrasound (LUS)

Thesis , University of Southern Denmark 2001. Author: Jesper Durup (please mail for reprint)

Summary

It is of great importance to know exactly the TNM stage in order to be able to design a correct therapy for patients with EC, GC and PC. In addition, a correct assessment of the resectability is a very important clinical factor, especially because futile laparotomies are important to avoid. Most imaging modalities have been too inaccurate in TNM staging and assessment of resectability. Other imaging techniques such as endoscopic ultrasound (EUS), seems to be accurate in certain cases, but are limited by stenosis and have problems in detecting dissemination of the cancer disease. Laparoscopy (LAP) is an accurate method in finding peritoneal carcinosis and hepatic metastases but has problems in detecting malignant lymph nodes and deep hepatic metastases. Furthermore LAP is not sufficient to assess the T-stage.

Laparoscopic Ultrasound (LUS) has in preliminary studies shown to be capable of solveing this problem, but the imaging technique have not been evaluated for TNM staging and assessment of resectability in esophageal cancer (EC), gastric cancer (GC) and pancreatic cancer (PC). The aim of this study was to investigate the efficacy of LUS in this respect.

LUS-guided biopsy has not been available before. The therapeutic possibilities and the possibility for increasing N and M-stage accuracy are important factors for introducing LUS-guided biopsies. During the study period, we have developed equipment for this purpose and started the evaluation. Although the results from the biopsy investigation are preliminary, they are promising.

The LUS probe is inserted through a laparoscopic trocar during the diagnostic laparoscopy. The tip is placed at the surface of the organ allowing the surgeon to investigate the whole organ. The probe is connected to an ultrasound unit, which transmits the ultrasound picture to the laparoscopic monitor. The study was a prospective, controlled, blinded investigation of 220 patients with EC, GC or PC. 113 patients were excluded due to disseminated disease, poor general condition or technical/administrative problems. Thus 107 patients were investigated with CT-scanning/abdominal ultrasound (CT/US), EUS, LAP and LUS. The gold standard was the histopathological evaluation of the resected specimen or biopsies from lymph nodes or metastases. In some cases explorative laparotomy or EUS, served as gold standard. The patients were staged and grouped according to the TNM definitions from UICC. Furthermore, they were grouped into three resectability groups. The LAP+LUS investigation was performed without knowledge of the result from the previous performed investigations. LUS was not able to T-stage esophageal cancer, but the accuracy in T-stage of GC and PC was high: 0.81 and 0.88 respectively, although there were only few T1 and T2 tumors. The N-stage accuracy was fair with accuracies of 0.68, 0.67 and 0.76 in EC, GC and PC respectively. The main problem in N staging was to discriminate between benign and malignant lymph nodes. The M staging was highly accurate when assessed by LUS with accuracies of 1.0, 0.88 and 0.85 in EC, GC and PC respectively. If the results gained from the LAP alone were added, the accuracies increased. LUS was highly accurate in the assessment of resectability when the patients were divided into three resectability groups. The accuracy for EC, GC and PC was 0.75, 0.80 and 0.80 respectively. When diagnosing non-resectable patients, the accuracies rose to 0.86, 0.98 and 0.91 respectively. In PC, the combination of different modalities were examined, in order to find the highest number of non-resectable patients. In the most accurate combinations of modalities, LAP+LUS were included.

There was a complication rate of 7%. The majority of the complications were minor, but three patients died postoperatively due to respiratory problems.

This study showed that LUS is capable of T staging of GC and CP, but is highly efficient in finding metastases and in the assessment of non-resectability in patients with upper GI cancer. LAP+LUS should be implemented in the staging algorithm in combination with EUS.