Self-Expanding Metallic Stents (SEMS) Versus Plastic Stents (PS) in Hilar Tumors: A Prospective Multicenter Cohort Study

Abstract

BACKGROUND: There is substantial controversy but virtually no comparative data as to whether SEMS or PS are optimal for drainage of hilar tumors. We tested the hypothesis that in patients with hilar tumors, SEMS result in fewer short term complications and repeat interventions than PS. METHODS: We performed a prospective observational cohort study examining outcomes of ERCP at 11 centers (6 private, 5 university). Data collected included detailed patient and procedural variables, and 30-day follow-up by a research nurse interview and chart review. Complications were defined by consensus criteria. Choice of SEMS or PS was per endoscopist and center preference. RESULTS: There were 62 cases (41 Bismuth II-IV) in which SEMS (N=34) or plastic stents (N=28) were placed in hilar tumors. Baseline characteristics were similar between stent groups with exceptions that compared with PS patients, SEMS patients more often had Bismuth III or IV (very complex) tumors (16/34 vs 5/28 p=0.043), higher Charlson comorbidity scores (mean 5.5 vs 3.7 p=0.003), and more frequent metastatic disease (19/34 vs 6/28 p=0.006). Bilateral stents were placed in 5/34 SEMS vs 4/28 PS (p=ns), with drainage assessed as incomplete in 6/34 SEMS vs 8/28 PS (p=ns). Adverse outcomes defined as cholangitis, stent malfunction (occlusion/perforation/migration), or unplanned percutaneous biliary drainage occurred in 4/34 (11.7%) of SEMS vs 11/28 (39.3%) PS (p=0.017). Multivariate predictors of adverse outcomes were placement of a plastic stent (OR 45.5, 3.6-500), Bismuth IV tumor (OR 38.8, 2.8-538), and bilirubin level (OR 1.15 per mg/dL above ULN, 1.0-1.3), but did not include Charlson comorbidity score, metastatic disease, endoscopists assessment of success or completeness of drainage, number of stents, or performance at the highest volume center. CONCLUSIONS: SEMS were markedly superior to plastic stents for drainage of hilar tumors with respect to short-term adverse events, independent of disease severity or quality of drainage. The high frequency of early adverse events with plastic stents suggests that preoperative plastic stenting should be avoided when possible and SEMS should be used preferentially in unresectable hilar tumors. A randomized comparative trial would be useful.