摘要
Background:Due to the clinically unapparent course the entity of left-sided pancreatic adenocarcinoma is often diagnosed at advanced stages,resulting in small numbers of patients qualifying for pancreatectomy.This study strives to develop a prognostic model for survival after left-sided pancreatic resection.Methods:A total of 54 patients were analyzed.Pre-and intra-operative predictive factors for 18-month mortality were identified with multivariable binary logistic regression analysis and compiled into a prognostic model.The applicability was evaluated by assessment of the area under the receiver operating characteristic curve(AUROC).The model was internally validated applying a randomized backwards bootstrapping analysis.Results:The 18-month mortality rate was 74.1%(n=40).Mean survival was 19.1 months.A prognostic model for 18-month mortality after left sided-pancreatectomy showed an AUROC>0.800:18-month mortality risk in%=Exp(Y)/(1+Exp(Y))with y=-0.927+(1.724,if CA 19-9 elevated,otherwise 0)+(1.212×number of intra-operative transfused packed red blood cells)+(2.771,if prior abdominal surgery,otherwise 0)-(3.612,if gastric resection,otherwise 0)This model was internally validated in 40 randomized backwards bootstrapping steps with AUROCs ranging from 0.757 to 0.971.Conclusions:The 18-month mortality risk for patients after left-sided pancreatectomy for adenocarcinoma of the pancreatic body can be assessed with the number of intra-operatively transfused packed red blood cells,elevated CA 19-9 levels,additional gastric resection and prior abdominal surgeries in the patient’s history.
Background: Due to the clinically unapparent course the entity of left-sided pancreatic adenocarcinoma is often diagnosed at advanced stages, resulting in small numbers of patients qualifying for pancreatectomy. This study strives to develop a prognostic model for survival after left-sided pancreatic resection. Methods: A total of 54 patients were analyzed. Pre-and intra-operative predictive factors for 18-month mortality were identified with multivariable binary logistic regression analysis and compiled into a prognostic model. The applicability was evaluated by assessment of the area under the receiver operating characteristic curve(AUROC). The model was internally validated applying a randomized backwards bootstrapping analysis. Results: The 18-month mortality rate was 74.1%( n = 40). Mean survival was 19.1 months. A prognostic model for 18-month mortality after left sided-pancreatectomy showed an AUROC > 0.800: 18-month mortality risk in% = Exp(Y)/(1 + Exp(Y)) with y =-0.927 +(1.724, if CA 19-9 elevated, otherwise 0) +(1.212 × number of intra-operative transfused packed red blood cells) +(2.771, if prior abdominal surgery, otherwise 0)-(3.612, if gastric resection, otherwise 0) This model was internally validated in 40 randomized backwards bootstrapping steps with AUROCs ranging from 0.757 to 0.971. Conclusions: The 18-month mortality risk for patients after left-sided pancreatectomy for adenocarcinoma of the pancreatic body can be assessed with the number of intra-operatively transfused packed red blood cells, elevated CA 19-9 levels, additional gastric resection and prior abdominal surgeries in the patient’s history.
