TY - JOUR
T1 - EUS diagnosis of vascular invasion in pancreatic cancer
T2 - Surgical and histologic correlates
AU - Aslanian, Harry
AU - Salem, Ronald
AU - Lee, Jeffrey
AU - Andersen, Dana
AU - Robert, Marie
AU - Topazian, Mark
PY - 2005/6/1
Y1 - 2005/6/1
N2 - BACKGROUND: Endoscopic ultrasound (EUS) has been compared to intraoperative surgical palpation for diagnosis of vascular invasion by pancreatic cancer. This study compares EUS with vascular resection and histologic evidence of vascular invasion in resected pancreatic masses. METHODS: All patients with solid pancreatic masses who underwent both preoperative EUS and surgery at 1 hospital over a 7 year period were identified. The relationship of pancreatic masses to adjacent vessels was prospectively assessed by EUS. EUS findings were compared to surgical and pathology gold standards. "Vascular adherence" was defined as tumor adherence requiring vascular resection during surgery, and "vascular invasion" as histologic invasion of vessel wall by tumor. RESULTS: 30 of 68 patients were resectable. Among these 30, vascular adherence was present in 8, including 18% of patients with an intact echoplane between tumor and adjacent vessels at EUS, 29% of those with loss of echoplane alone, and 50% of those with additional EUS features of vascular involvement. Vascular invasion was present in 4, including 12% of patients with an intact echoplane, 0% of those with loss of echoplane alone, and 33% of those with additional EUS features. Sensitivity, specificity, PPV, and NPV of EUS were 63%, 64%, 43% and 80% for vascular adherence and 50% 58%, 28% and 82% for vascular invasion. NPV rose to 90% for vascular adherence if only the portal confluence vessels were considered. CONCLUSIONS: EUS has poor sensitivity, specificity, and positive predictive value for diagnosis of venous involvement by pancreatic cancer.
AB - BACKGROUND: Endoscopic ultrasound (EUS) has been compared to intraoperative surgical palpation for diagnosis of vascular invasion by pancreatic cancer. This study compares EUS with vascular resection and histologic evidence of vascular invasion in resected pancreatic masses. METHODS: All patients with solid pancreatic masses who underwent both preoperative EUS and surgery at 1 hospital over a 7 year period were identified. The relationship of pancreatic masses to adjacent vessels was prospectively assessed by EUS. EUS findings were compared to surgical and pathology gold standards. "Vascular adherence" was defined as tumor adherence requiring vascular resection during surgery, and "vascular invasion" as histologic invasion of vessel wall by tumor. RESULTS: 30 of 68 patients were resectable. Among these 30, vascular adherence was present in 8, including 18% of patients with an intact echoplane between tumor and adjacent vessels at EUS, 29% of those with loss of echoplane alone, and 50% of those with additional EUS features of vascular involvement. Vascular invasion was present in 4, including 12% of patients with an intact echoplane, 0% of those with loss of echoplane alone, and 33% of those with additional EUS features. Sensitivity, specificity, PPV, and NPV of EUS were 63%, 64%, 43% and 80% for vascular adherence and 50% 58%, 28% and 82% for vascular invasion. NPV rose to 90% for vascular adherence if only the portal confluence vessels were considered. CONCLUSIONS: EUS has poor sensitivity, specificity, and positive predictive value for diagnosis of venous involvement by pancreatic cancer.
UR - http://www.scopus.com/inward/record.url?scp=21344454002&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=21344454002&partnerID=8YFLogxK
U2 - 10.1111/j.1572-0241.2005.41675.x
DO - 10.1111/j.1572-0241.2005.41675.x
M3 - Article
C2 - 15929774
AN - SCOPUS:21344454002
SN - 0002-9270
VL - 100
SP - 1381
EP - 1385
JO - American Journal of Gastroenterology
JF - American Journal of Gastroenterology
IS - 6
ER -