The pancreaticojejunal anastomotic stent: Friend or foe?

Background. The efficacy of pancreaticojejunal (P-J) anastomotic stents in preventing clinically relevant postoperative pancreatic fistulas (CR-POPF) after pancreatic resection is poorly understood. We sought to compare the outcomes of stented and nonstented patients in light of recognized risk-factors for the development of CR-POPF and to determine whether outcomes differed once there was a change in practice where use of stents was abandoned. Methods. A total of 444 patients underwent proximal pancreatic resection with P-J reconstruction from 2001 to 2011. At the surgeon's discretion, a PJ stent (5- or 8-Fr Silastic tube) was placed in 59 patients (13.3%; 46 internal, 13 external). Demographics, comorbidities, and adjusted outcomes were evaluated between groups of nonstented (n = 385) and stented patients; these outcomes included a subgroup analysis of internally and externally stented patients. Risk factors for CR-POPF (International Study Group on Pancreatic Fistula grade B/C) development have been previously defined as soft gland, small duct size, high-risk pathology, or excessive blood loss (>1,000 mL). Outcomes were interpreted in reference to the risk factor profile (the number of absolute risk factors present; 0-4), and to the fistula risk score, a prospectively validated score which accurately predicts the risk and impact of pancreatic fistula based on these variables. Results. Preoperative demographics of age, sex, body mass index, American Society of Anesthesiologists class, and physiologic and operative severity score for the enumeration of mortality and morbidity (ie, POSSUM) score were equivalent between cohorts. The CR-POPF risk-factor profile and fistula risk score were greater in stented patients (P < .01). When compared with nonstented patients, stented patients actually had greater rates of CR-POPF (29% vs 11%), major complications (29% vs 14%), greater mean duration of stay (13.7 days vs 9.6 days), and total costs ($33,594 vs $22,411; all P < .05). When high-risk cases were scrutinized, P-J stent use did not offer protection, as CR-POPF was uniformly more common when stents were used. Rates and severity of CR-POPF did not increase when the use of stents was abandoned, further implying that they did not confer protection from fistula development. Extended postoperative imaging was available for 23 stented patients. Of these, one-third of stents were retained past 6 weeks, and one-fourth beyond 6 months. Four patients required additional procedures to manage stent-related complications. Conclusion. The use of P-J stents does not decrease the incidence or severity of CR-POPF after proximal pancreatic resection, both overall and for high-risk scenarios. In some patients, PJ stents may lead to short- and long-term adverse outcomes. (Surgery 2013;153:651-62.)