A further modification of the blumgart pancreatojejunostomy: results of a propensity score-matched analysis versus cattel-warren pancreatojejunostomy in open and robotic pancreatoduodenectomy

Abstract

Introduction: Appropriate surgical technique is key to reduce incidence and severity of post-operative pancreatic fistula (POPF). Method: Blumgart pancreatojejunostomy was further modified (m-BPJ), by reducing to 2 the number of transparenchimal sutures and by adding two two “half purse-string sutures” at the corners. m-BPJ was compared to Cattell-Warren pancreatojejunostomy (C-WPJ) before and after propensity score matching in both open (OPD) and robotic (RPD) pancreatoduodenectomy. The primary study endpoint was incidence of clinically relevant POPF (CR-POPF). Results: mBPJ was used in 190 patients (124 OPD; 66 RPD). C-WPJ was employed in 225 patients (143 OPD and 82 RPD). The incidence of CR-POPF was 13.7% and 29.8% (p< 0.0001; OR=0.37) in mBPJ and C-WPJ, respectively. Equivalent figures in OPD and RPD subgroups were 13.7% and 28.8% (p< 0.003; OR= 0.40) and 13.6% and 31.7% (p< 0.01; OR= 0.34), respectively. Grade C POPF occurred in 9 patients after C-WPJ (4.0%) and in 1 patient after mBPJ (0.5%) (p=0.02). Predictors of CR-POPF are reported in table 1. The matching process identified 107 pairs (78 OPD and 29 RPD). In the global population the adjusted OR decreased to 0.40 (p=0.0007). CR-POPF occurred in 12.8% and 28.2% (p=0.01) OPDs, and in 6.9% and 41.4% (p=0.002) RPDs using mBPJ and C-WPJ, respectively. In propensity score matched analysis OR for CR-POPF of mBPJ decreased to 0.36 (p=0.18) and to 0.09 (p=0.01) in OPDs and RPDs, respectively. Conclusion: In our analysis mBPJ decreased incidence of CR-POPF as compared to C-WPJ in the general population, in OPD and RPD.Tabled 1[Table 1 - Predictive factors of CR-POPF].