Endoscopic stenting for inoperable malignant biliary obstruction: A systematic review and meta-analysis.

AIM: To analyze through meta-analyses the benefits of two types of stents in the inoperable malignant biliary obstruction.

METHODS: A systematic review of randomized clinical trials (RCT) was conducted, with the last update on March 2015, using EMBASE, CINAHL (EBSCO), MEDLINE, LILACS/CENTRAL (BVS), SCOPUS, CAPES (Brazil), and gray literature. Information of the selected studies was extracted in sight of six outcomes: primarily regarding dysfunction, complication and re-intervention rates; and secondarily costs, survival, and patency time. The data about characteristics of trial participants, inclusion and exclusion criteria and types of stents were also extracted. The bias was mainly assessed through the JADAD scale. This meta-analysis was registered in the PROSPERO database by the number CRD42014015078. The analysis of the absolute risk of the outcomes was performed using the software RevMan, by computing risk differences (RD) of dichotomous variables and mean differences (MD) of continuous variables. Data on RD and MD for each primary outcome were calculated using the Mantel-Haenszel test and inconsistency was qualified and reported in χ (2) and the Higgins method (I (2)). Sensitivity analysis was performed when heterogeneity was higher than 50%, a subsequent assay was done and other findings were compiled. Student's t-test was used for the comparison of weighted arithmetic means regarding secondary outcomes.

RESULTS: Initial searching identified 3660 studies; 3539 were excluded through title, repetition, and/or abstract, while 121 studies were fully assessed and were excluded mainly because they did not compare self-expanding metal stents (SEMS) and plastic stents (PS), leading to thirteen RCT selected, with 13 articles and 1133 subjects meta-analyzed. The mean age was 69.5 years old, that were affected mostly by bile duct (proximal) and pancreatic tumors (distal). The preferred SEMS diameter used was the 10 mm (30 Fr) and the preferred PS diameter used was 10 Fr. In the meta-analysis, SEMS had lower overall stent dysfunction compared to PS (21.6% vs 46.8%, P < 0.00001) and fewer re-interventions (21.6% vs 56.6%, P < 0.00001), with no difference in complications (13.7% vs 15.9%, P = 0.16). In the secondary analysis, the mean survival rate was higher in the SEMS group (182 d vs 150 d, P < 0.0001), with a higher patency period (250 d vs 124 d, P < 0.0001) and a lower cost per patient (4193.98 vs 4728.65 Euros, P < 0.0985).

CONCLUSION: SEMS are associated with lower stent dysfunction, lower re-intervention rates, better survival, and higher patency time. Complications and costs showed no difference.