Which Type of Stent Is Useful for Malignant Gastrointestinal Obstruction?: In Viewpoint of Mechanism of Obstruction

Abstract

Which Type of Stent Is Useful for Malignant Gastrointestinal Obstruction?: In Viewpoint of Mechanism of Obstruction Deog K. Kim, Jai H. Jeung, Choong H. Cha, Jae Y. Cheong, Kee M. Lee, Byung M. Yoo, Kwang J. Lee, Ki B. Hahm, Jin H. Kim, Sung W. Cho Background: Malignant upper gastrointestinal(GI) obstruction is the late complication of gastric, pancreatic, bile duct, and gallbladder cancer. The mechanism of malignant GI obstruction could be categorized into two basic types; endogenous direct tumor growth into the lumen and exogenously extrinsic compression/invasion of the tumor. In this study, we compared the feasibility and outcome of covered stents (CS) and uncovered stents (US) in malignant GI obstruction according to the mechanism of obstruction. Methods: Between January 1998 and October 2005, 157 patients (108 men, 49 women: mean age 65.2) presenting upper GI obstruction with unresectable malignancy underwent endoscopic stent insertion. Complications and clinical outcome were assessed according to the stent type and the mechanism of obstruction. Results: 1) All of 131 patients with the endogenous GI obstruction (advanced gastric cancer 126, ampullary cancer 4, and duodenal cancer 1) were treated with the insertion of CS (n Z 50) and US (n Z 81). All of 26 patients with the exogenous GI obstruction (pancreatic cancer 17, bile duct cancer 3, GB cancer 3, colon cancer 2 and ovarian cancer 1) were treated with the insertion of CS (n Z 9) and US (n Z 17). 2) Clinical success rates of CS and US insertion were 98% (49/50) and 96.2% (80/81) in the endogenous obstruction, 100% (9/9) and 100% (17/17) in the exogenous obstruction. 3) Overall complications of CS and US were tumor ingrowth (5.2%, 20.0%; p Z 0.011), overgrowth (5.2%, 2.1%; p O 0.05) and migration (20.7%, 1.1%; p ! 0.001). 4) The complications of CS and US in the endogenous GI obstruction were ingrowth (6.1%, 19.2%; p Z 0.039), overgrowth (6.1%, 2.6%; p O 0.05) and migration (17.5%, 1.3%; p ! 0.001). 5) The complications of CS and US in the exogenous GI obstruction were ingrowth (0%, 23.5%; p O 0.05), overgrowth (0%, 0%) and migration (22.23%, 0%; p O 0.05). 6) The patency duration of CS in the endogenous GI obstruction was significantly longer than that of US (129 G 109.9, 89 G 82.3; p Z 0.036). There was no statistically significant difference in the patency duration between CS and US in the exogenous GI obstruction (63 G 70.3, 63 G 57.4; p Z 0.99). Conclusion: Covered stents would be more feasible for the malignant gastrointestinal obstruction induced by the endogenous stenosis because of the longer stent duration.