Potential significance of uncovered self-expandable metal stents for distal malignant biliary obstruction: A propensity score-adjusted competing risk regression analysis.

Objectives: Selection criteria for self-expandable metal stents (SEMSs) with or without cover during palliative treatment of distal malignant biliary obstruction (DMBO) remain unclear. We evaluated factors associated with time to recurrent biliary obstruction (TRBO) in fully covered SEMSs (FCSEMSs) and uncovered SEMSs (UCSEMSs). Methods: We retrospectively analyzed consecutive patients with DMBO who received a SEMS. TRBO was determined using the Kaplan-Meier analysis, and complications were compared between the FCSEMS and UCSEMS groups. After TRBO-associated factors were extracted using multivariate competing-risks regression (CRR), propensity score-adjusted CRRs were performed to verify their robustness. Results: There were 180 patients (66 FCSEMSs and 114 UCSEMSs) enrolled in this study. There was no significant difference between median TRBO in the FCSEMS and UCSEMS groups (275 vs. 255 days, p = 0.67). Complications were more frequent in the FCSEMS than UCSEMS group (21.2% vs. 8.8%; p = 0.023). Multivariate CRR for TRBO-associated factors revealed that "pancreatic ductal carcinoma (PDAC) treated with UCSEMS" was the only independent predictor of TRBO (p = 0.03). Similarly, the propensity score-adjusted CRRs showed no significant difference in TRBO in "FCSEMS" vs "UCSEMS" (p = 0.96); however, there was a significant difference in "PDAC using UCSEMS" vs "other" (p = 0.043). In the palliative care group including any DMBO without chemotherapy, the first quartile of the TRBO of UCSEMS was 100 days. Conclusions: UCSEMSs are a possible option for both patients with DMBO arising from PDAC and for patients with any DMBO receiving palliative care who should avoid SEMS-related complications.