Medline ® Abstract for Reference 12
of 'Endoscopic ultrasound in the staging of exocrine pancreatic cancer'
Role of EUS in the preoperative staging of pancreatic cancer: a large single-center experience.
Gress FG, Hawes RH, Savides TJ, Ikenberry SO, Cummings O, Kopecky K, Sherman S, Wiersema M, Lehman GA
Gastrointest Endosc. 1999;50(6):786.
BACKGROUND: Current methods for staging pancreatic cancer can be inaccurate, invasive, and expensive. Endoscopic ultrasound (EUS) is reported to be highly accurate for local staging of gastrointestinal tumors including pancreatic cancer. The aim of this study was to assess the utility of EUS and CT for staging pancreatic cancer by comparing staging accuracies in surgical patients and evaluating the potential impact of EUS staging and training.
METHODS: This was a preoperative comparison of the diagnostic operating characteristics of these procedures in a referral-based academic medical center. Data were collected on 151 consecutive patients referred with confirmed pancreatic cancer between April 1990 and November 1996. All patients had preoperative CT and EUS performed for staging. In patients undergoing surgery, the surgical staging and/or findings were used to confirm EUS and CT staging.
RESULTS: Eighty-one (60%) of 151 patients underwent surgery and made up the study subset. In these 81 patients, surgical exploration provided a final T staging in 93% (75 of 81), N staging in 88% (71 of 81) and data on vascular invasion in 93% (75 of 81). In the surgical patient group, with surgical correlation, EUS accuracy for T staging was as follows: T1 92%, T2 85%, T3 93%, and for N staging was: N0 72%, and N1 72%. CT accuracy for T staging was as follows: T1 65%, T2 67%, T3 38%, and for N staging was as follows: N0 52% and N1 100%. CT failed to detect a mass in 26% of patients with a confirmed tumor at surgery. Overall accuracy for T and N staging was 85% and 72% for EUS and 30% and 55% for CT, respectively. The ability to accurately predict vascular invasion was 93% for EUS and 62% for CT (p<0.001). EUS was 93% accurate for predicting local resectability versus 60% for CT (p<0.001). Last, the data were divided into two groups for the senior endosonographer's experience: procedures performed between 1990 and 1992 (98 cases) and 1993 and 1994 (53 cases). This analysis revealed that 7 of 9 instances of mis-staging (78%) occurred in the earlier group, during the learning phase for EUS.
CONCLUSIONS: EUS is more accurate than CT for staging pancreatic malignancies, including predicting vascular invasion and local resectability. EUS staging was significantly better than CT for T1, T2, and T3 tumors. EUS staging accuracy improved after 100 cases, thus suggesting a correlation between the accuracy of EUS staging and the number of procedures performed.
Division of Gastroenterology and Hepatology, Departments of Pathology and Radiology, Indiana University School of Medicine, Indiana University Medical Center, Indianapolis, Indiana, USA.