Search

Background:Postpancreaticoduodenectomy (PD) hemorrhage (PPH) is a life-threatening complication after PD. The main objective of this study was to evaluate incidence and factors associated with late PPH as well as the management strategy and outcomes. Methods: Between May 2017 and March 2020, clinical data from 192 patients undergoing PD were collected prospectively in the CHIRPAN Database (NCT02871336) and retrospectively analyzed. In our institution, all patients scheduled for a PD are routinely admitted for monitoring and management in intensive/intermediate care unit (ICU/IMC). Results: The incidence of late PPH was 17% (32 of 192), whereas the 90-day mortality rate of late PPH was 19% (6 of 32). Late PPH was associated with 90-day mortality (P = 0.001). Using multivariate analysis, independent risk factors for late PPH were postoperative sepsis (P = 0.036), and on day 3, creatinine (P = 0.025), drain fluid amylase concentration (P = 0.023), lipase concentration (P0.001), and C-reactive protein (CRP) concentration (P0.001). We developed two predictive scores for PPH occurrence, the PANCRHEMO scores. Score 1 was associated with 68.8% sensitivity, 85.6% specificity, 48.8% predictive positive value, 93.2% negative predictive value, and an area under the receiver operating characteristic curves of 0.841. Score 2 was associated with 81.2% sensitivity, 76.9% specificity, 41.3% predictive positive value, 95.3% negative predictive value, and an area under the receiver operating characteristic curve of 0.859. Conclusions: Routine ICU/IMC monitoring might contribute to a better management of these complications. Some predicting factors such as postoperative sepsis and biological markers on day 3 should help physicians to determine patients requiring a prolonged ICU/IMC monitoring.

Introduction The optimal management of rectal cancer with synchronous liver metastases remains debatable. Thus, we propose an optimised liver-first (OLF) strategy that combines concomitant pelvic irradiation with hepatic management. This study aimed to evaluate the feasibility and oncological quality of the OLF strategy. Materials and methods Patients underwent systemic neoadjuvant chemotherapy followed by preoperative radiotherapy. Liver resection was performed in one step (between radiotherapy and rectal surgery) or in two steps (before and after radiotherapy). The data were collected prospectively and analysed retrospectively as intent to treat. Results Between 2008 and 2018, 24 patients underwent the OLF strategy. The rate of treatment completion was 87.5%. Three patients (12.5%) did not proceed to the planned second-stage liver and rectal surgery because of progressive disease. The postoperative mortality rate was 0%, and the overall morbidity rates after liver and rectal surgeries were 21% and 28.6%, respectively. Only two patients developed severe complications. Liver and rectal complete resection was performed in 100% and 84.6%, respectively. A rectal-sparing strategy was performed in 6 patients who underwent local excision (n = 4) or a watch and wait strategy (n = 2). Among patients who completed treatment, the median overall and disease-free survivals were 60 months (range 12–139 months) and 40 months (range 10–139 months), respectively. Eleven patients (47.6%) developed recurrence, among whom five underwent further treatment with curative intent. Conclusion The OLF approach is feasible, relevant, and safe. Organ preservation was feasible for a quarter of patients and may be associated with reduced morbidity.

Background Factors excluding postoperative pancreatic fistula (POPF), facilitating early drain removal and hospital discharge represent a novel approach in patients undergoing enhanced recovery after pancreatic surgery. This study aimed to establish the relevance of neutrophil-to-lymphocyte ratio (NLR) in excluding POPF after pancreatoduodenectomy (PD). Methods A prospectively maintained database of patients who underwent PD at two high-volume centres was used. Patients were divided into three cohorts (training, internal, and external validation). The primary endpoints of this study were accuracy, optimal timing, and cutoff values of NLR for excluding POPF after PD. Results From 2012 to 2020, in a 2:1 ratio, 451 consecutive patients were randomly sampled as training (n = 301) and validation (n = 150) cohorts. Additionally, the external validation cohort included 197 patients between 2018 and 2020. POPF was diagnosed in 135 (20.8 per cent) patients. The 90-day mortality rate was 4.1 per cent. NLR less than 8.5 on postoperative day 3 (OR, 95 per cent c.i.) was significantly associated with the absence of POPF in the training (2.41, 1.19 to 4.88; P = 0.015), internal validation (5.59, 2.02 to 15.43; P = 0.001), and external validation (5.13, 1.67 to 15.76; P = 0.004) cohorts when adjusted for relevant clinical factors. Postoperative outcomes significantly differed using this threshold. Conclusion NLR less than 8.5 on postoperative day 3 may be a simple, independent, cost-effective, and easy-to-use criterion for excluding POPF.

PURPOSE The double purse-string telescoped pancreaticogastrostomy (PG) technique has been suggested as an alternative approach to reduce the risk of postoperative pancreatic fistula (POPF). Its efficacity in high-risk situations has not yet been explored. This study compared the incidence of clinically relevant POPF (CR-POPF) between patients with high-risk anastomosis undergoing PG and those undergoing pancreaticojejunostomy (PJ). METHODS From 2013 to 2019, 198 consecutive patients with high-risk anastomosis, an updated alternative fistula risk score > 20%, and who underwent pancreatoduodenectomy with the PJ (165) or PG (33) technique were included. Optimal mitigation strategy (external stenting/octreotide omission) was applied for all patients. The primary endpoint was the incidence of CR-POPF. RESULTS The mean ua-FRS was 33%. CR-POPF (grade B/C) was found in 42 patients (21%) and postoperative hemorrhage in 30 (15%); the mortality rate was 4%. CR-POPF rates were comparable between the PJ (19%) and PG (33%) groups (P = 0.062). The PG group had a higher rate of POPF grade C (24% vs. 10%; P = 0.036), longer operative time (P = 0.019), and a higher transfusion rate (P

Several scoring systems predict risks of clinically relevant postoperative pancreatic fistula after pancreatectomy, but none have emerged as the gold standard. This study aimed to evaluate the accuracy of postoperative day 1 drain amylase and serum C-reactive protein levels in predicting clinically relevant postoperative pancreatic fistula compared with intraoperative pancreatic characteristics.Patients who underwent pancreatectomy between 2017 and 2019 were included prospectively. Cutoff values were determined using receiver operating characteristic curves, and a score combining postoperative day 1 drain amylase and serum C-reactive protein was tested in a multivariate logistic regression model to evaluate clinically relevant postoperative pancreatic fistula risk.A total of 274 pancreatic resections (182 pancreaticoduodenectomies and 92 distal pancreatectomies) were included. The pancreatic gland texture was "soft" in 47.8% (n = 131), and 55.8% (n = 153) had a small size main pancreatic duct (≤3 mm). Clinically relevant postoperative pancreatic fistula occurred in 58 patients (21.2%). Drain amylase ≥1,000 UI/L and serum C-reactive protein ≥90 mg/L were identified as the optimal cutoffs to predict clinically relevant postoperative pancreatic fistula. On multivariate analysis these cutoffs were independent predictors of clinically relevant postoperative pancreatic fistula after both pancreaticoduodenectomies (drain amylase: P.001, serum C-reactive protein: P = .006) and distal pancreatectomies (drain amylase: P = .009, serum C-reactive protein: P = .001). The postoperative day 1 "90-1000" model, a 2-value score relying on these cutoffs, significantly (P.001) outperformed intraoperative pancreatic parenchymal characteristics in predicting clinically relevant postoperative pancreatic fistula after both pancreaticoduodenectomies and distal pancreatectomies. A postoperative day 1 "90-1000" score = 0 had a negative predictive value of 97% and 94%, respectively, after pancreaticoduodenectomy and distal pancreatectomies.A combined score relying on postoperative day 1 values of drain amylase and serum C-reactive protein levels was accurate in predicting risks of clinically relevant postoperative pancreatic fistula after pancreatectomy.

Background Emergency completion pancreatectomy (CP) after pancreatoduodenectomy (PD) is a technically demanding procedure. We report our experiences with a four-step standardized technique used at our center since 2012. Methods In the first step, the gastrojejunostomy is divided with a stapler to quickly access the pancreatic anastomosis and permit adequate exposure, especially in cases of active bleeding. Second, the bowel loops connected to the pancreatic anastomosis is divided in cases of pancreaticojejunostomy. Third, the pancreatectomy is completed with or without the splenic vessels and spleen conservation according to the local conditions. Finally, the fourth step reconstructs in a Roux-en-Y fashion and ensures drainage. Results From January 2012 to December 2019, 450 patients underwent PD at our center. Reintervention for grade C postoperative pancreatic fistula was decided for 30 patients, and CP was performed in 21 patients. The mean intraoperative blood loss and operative duration were relatively low (600 ml and 240 min, respectively). During the perioperative period, three patients died from multiple organ failure, and two patients died intraoperatively from a cataclysmic hemorrhage originating from the superior mesenteric artery. Discussion Our standardized procedure appears to be relatively safe, reproducible, and could be particularly useful for young surgeons.

The Birmingham score predicts the risk of hospital readmission after pancreaticoduodenectomy (PD). This study aimed to validate the risk score in a different healthcare cohort.From 2017 to 2021, 301 patients underwent PD. The Birmingham score was applied to 276 patients. Postoperative deceased patients (n = 7) or those requiring a completion of pancreatectomy (n = 18) were excluded.Forty-seven (17%) patients were readmitted after a median delay of 9 (range 1-49) days and stayed for 5 (range 1-27) days; 4 (8.5%) died during the hospital stay. The leading cause of readmission was a septic condition (53%), mostly resolved by medical treatment (77%). A multivariate analysis identified the occurrence of a clinically relevant postoperative pancreatic fistula, the score criteria, and the score itself as independent factors favouring readmission. Readmission rates in patients with low [n = 97 (35%)], intermediate [n = 98 (36%)], and high [n = 81 (29%)] scores were 5%, 17%, and 31%, respectively (P 0.01).This study confirmed the relevance and robustness of the Birmingham risk score. Patients with a high risk of readmission after PD, identified based on the score, were discharged to a partnership medical centre close to the pancreatic centre to plan readmission and avoid futile unplanned hospitalisation.

International audience; Background: Experienced pancreatic surgeons, for whom complexity is not an issue, must decide at the end of neoadjuvant therapy whether to continue or discontinue surgery, when pancreatectomy with vascular resection is planned in patients with pancreatic ductal adenocarcinoma (PDAC). Objective: Our study aimed to determine preoperative factors that can predict short postoperative survival in such situations. Methods: Overall, 105 patients with borderline or locally advanced PDAC received neoadjuvant FOLFIRINOX (followed by chemoradiation in 22% of patients) and underwent pancreatectomy with segmental venous and/or arterial resection at two high-volume centers. The primary endpoint was overall survival (OS) of < 1 year after surgery for patients who did not die from the surgery. Results: Tumors were classified as borderline in 78% of cases and locally advanced in 22% of cases. Mean CA19-9 at diagnosis was 934 U/mL, which significantly decreased to 213 U/mL (p < 0.01) after a median of six cycles of FOLFIRINOX. Pancreaticoduodenectomy was performed most often (76%). The vast majority of patients underwent venous resection (92%), and a simultaneous arterial resection was performed in 16 patients (15%). The severe morbidity rate and 30- and 90-day mortality rates were 21%, 8.5%, and 10.4%, respectively. The median OS after surgery was 23 months. In the multivariate analysis, preoperative CA19-9 ≥ 450 U/mL was the only preoperative factor independently associated with OS of < 1 year (p = 0.044). Conclusion: The preoperative CA19-9 value should be considered in the clinical decision-making process when complex vascular resection is required. © 2021, Society of Surgical Oncology.

Malignant insulinomas are functional neuroendocrine tumors of the pancreas and the primary cause of tumor-related hypoglycemia. Malignant insulinoma is rare and has a poor prognosis. We report a case of metastatic malignant insulinoma in a 64-year-old female patient with severe and refractory hypoglycemia. After several ineffective locoregional and systemic therapeutic lines for the secretory disease, the introduction of pasireotide, a second-generation somatostatin analog, provided an improved clinical and secretory evolution both quickly and sustainably, with an excellent safety profile. Pasireotide is an effective and well-tolerated therapy in the treatment of refractory hypoglycemia in metastatic insulinoma.

An amendment to this paper has been published and can be accessed via the original article.

This study aimed to determine the impact of FOLFIRINOX neoadjuvant therapy on patients with non-metastatic borderline/locally advanced (BL/LA) pancreatic ductal adenocarcinoma (PDAC), in current practice.From 2010 to 2017, 258 patients with BL/LA PDAC from a single high-volume institution received FOLFIRINOX neoadjuvant treatment.The 258 patients received a median number of 6 cycles of FOLFIRINOX (range, 3-16); 98 (38%) patients underwent curative surgery, and 160 (62%) continued medical treatment. A venous resection was performed in 57 patients (58%), and an arterial resection in 12 (12%). The postoperative 30- and 90-day mortality rates were 6.1% and 8.2%, respectively. Adjuvant chemotherapy was performed in 57 patients (59%). The median overall survival (OS) in patients who did (n = 98) or did not (n = 160) undergo surgical resection were 39 months and 19 months, respectively (P 0.001). In resected patients, the ASA 3 score (P 0.01), venous resection (P 0.01), hemorrhage (P 0.01), and R1 margin status (P = 0.03) were found to negatively influence the OS. The median OS was significantly higher in patients who did not require a venous resection (not reached vs. 26.5 months, P 0.001).Neoadjuvant FOLFIRINOX provided a survival benefit in BL/LA PDAC patients, particularly in those who did not ultimately require venous resection.

We evaluated the intuition of expert pancreatic surgeons, in predicting the associated risk of pancreatic resection and compared this “intuition” to actual operative follow-up. The objective was to avoid major complications following pancreatic resection, which remains a challenge. From January 2015 to February 2018, all patients who were 18 years old or more undergoing a pancreatic resection (pancreaticoduodenectomy [PD], distal pancreatectomy [DP], or central pancreatectomy [CP]) for pancreatic lesions were included. Preoperatively and postoperatively, all surgeons completed a form assessing the expected potential occurrence of clinically relevant postoperative pancreatic fistula (CR-POPF: grade B or C), postoperative hemorrhage, and length of stay. Preoperative intuition was assessed for 101 patients for 52 PD, 44 DP, and 5 CP cases. Overall mortality and morbidity rates were 6.9% (n = 7) and 67.3% (n = 68), respectively, and 38 patients (37.6%) developed a POPF, including 27 (26.7%) CR-POPF. Concordance between preoperative intuition of CR-POPF occurrence and reality was minimal, with a Cohen’s kappa coefficient (κ) of 0.175 (P value = 0.009), and the same result was obtained between postoperative intuition and reality (κ = 0.351; P

We evaluated the usefulness of the 2017 definition of borderline pancreatic ductal adenocarcinoma (BR-PDAC) in fit patients (performance status 0-1) based on anatomical (A) and biological dimensions (B).From 2011 to 2018, 139 resected patients with BR-PDAC according to the 2017 definition were included: 18 patients underwent upfront pancreatectomy (CA 19-9 500 U/mL and/or regional lymph node metastasis; BR-B group), and 121 received FOLFIRINOX (FX) induction chemotherapy and were divided into BR-A (CA 19-9 500 U/mL, no regional lymph node metastasis; n = 68) and BR-AB (CA 19-9 500 U/mL and/or regional lymph node metastasis; n = 53) groups.The 3 groups were comparable according to patient characteristics (except for back pain (P .01) and CA 19-9 (P .01)), intraoperative data, and postoperative courses. BR-AB patients required more venous resections (P .01). The 3 groups were comparable on pathologic findings, except that BR-B patients had more lymph node invasions (P = .02). Median overall survival (OS) of the 121 patients was 45 months. In multivariate analysis, venous resection (P = .039) and R1 resection (P = .012) were poorly linked with OS, whereas BR-A classification (P .01) independently favored OS. Median survival times of BR-A, BR-AB, and BR-B groups were undetermined, 27 months, and 20 months (P.001), respectively.The 2017 definition was relevant for sub-classifying patients with BR-PDAC. The anatomical dimension (BR-A) was a favorable prognostic factor, whereas the biological dimension (BR-AB and BR-B) poorly impacted survival.

Two-stage hepatectomy (TSH) strategy is used to treat patients with bilobar colorectal liver metastasis (CLM). However, many patients do not undergo the second hepatectomy owing to disease progression in the future liver remnant (FLR) after portal vein embolization (PVE). This study aimed to assess the impact of disease progression in the FLRs of patients who completed the first hepatectomy.68 consecutive patients underwent the first hepatectomy followed by PVE. Six patients (9%) dropped out after the PVE (two-stage failed [TSF] group) because of unresectable hepatic or general disease progression. Seventeen patients (25%) completed their second hepatectomy despite disease progression in the FLR (new CLM [nCLM] group) as it was considered resectable, while 45 patients (66%) underwent the second hepatectomy (control group).The 5-year overall survival rates in the TSF, nCLM, and control groups were 0%, 7%, and 60%, respectively (P 0.001). The median overall survival times between the TSF and nCLM groups were 26 months and 42 months (P = 0.005). Patients in the nCLM group whose hepatic disease progression was detected preoperatively versus intraoperatively had comparable survival rates.Resectable hepatic disease progression in the FLR after PVE should not be considered a contraindication for the second hepatectomy.

Surgical field photography is a tough exercise: surgeons dedicate the required time for photography even during complex surgeries; the intense lighting of the operating field works against photography, and the surgeon has to utilise whatever equipment is available. We selected five complex interventions and two surgeons (one with an iPhone

Summary Purpose Survival appears to be poor in cases of pancreatic ductal adenocarcinoma (PDAC) with para-aortic lymph node involvement (PALN+). However, resection is still performed in these cases because the prognostic impact of PALN+ remains controversial. Methods PALN+ was intraoperatively found in 14 patients (4.8%) with resectable PDAC who consequently did not undergo pancreatectomy. Results The median overall survival time after laparotomy was 21 months. The 1- and 3-year overall survival rates were 58.3% and 25%, respectively. Conclusions We support the advisability of reconsidering pancreatectomy in patients with intraoperatively detected PALN+ because the reported survival of such patients who undergo pancreatectomy is poorer than the survival observed for patients in our series.

Resume Introduction La survie des patients presentant un adenocarcinome pancreatique canalaire infiltrant est limitee en cas d’envahissement ganglionnaire para-aortique. La resection pancreatique est toutefois encore souvent proposee du fait d’un impact pronostique qui demeure debattu. Methode Un envahissement ganglionnaire para-aortique etait retrouve en peroperatoire chez 14 patients (4,8 %) presentant un adenocarcinome pancreatique canalaire infiltrant resecable et chez qui la pancreatectomie n’a pas ete realisee. Resultats La mediane de survie globale apres laparotomie etait de 21 mois. Les survies a 1 an et 3 ans etaient respectivement de 58 et 25 %. Conclusions Nous discutons le bien-fonde d’un geste de resection pancreatique en cas d’envahissement ganglionnaire para-aortique car les survies reportees chez les patients reseques sont inferieures a la survie des patients de notre serie.

To describe, in patients with resectable pancreatic ductal adenocarcinoma (PDAC), the laparotomy findings, treatments and outcomes before (period 1) and after 2010 (period 2).From 2000 to 2015, patients newly diagnosed with resectable PDAC at Paoli-Calmettes Institute, France, were evaluated. Survival was examined using the Kaplan-Meier method, and statistical comparisons were conducted using log rank tests.Among 1175 patients diagnosed with pancreatic mass, 164 underwent laparotomy with an intention of pancreatic resection. Some of them did not undergo pancreatic resection due to peroperative discovery of advanced disease. For those who were finally resected (n = 119), there were fewer pancreaticoduodenectomies (p = 0.045), shorter operation times (p 0.01), lower mortality rates (p = 0.02), more advanced-stage tumors (T3), more frequent perineural invasion and R1 resection in period 2. This group had a trend of better outcomes after 2010 (51 months vs. 36 months (p = 0.065)).Improvement in surgical procedures and postoperative management led to prolonged survival of those who underwent surgery for resectable pancreatic cancer since 2010, despite a higher frequency of advanced tumors at the diagnosis in our institution.

Positive para-aortic lymph nodes (PALN) (station 16) are commonly detected in the final pathologic examination (ranging from 15 to 26%) among patients who undergo upfront pancreatoduodenectomy for resectable pancreatic ductal adenocarcinoma. However, after neoadjuvant treatment (NAT) the role of positive PALN as a watershed for surgical resection remains unclear. We aimed to determine the incidence of intraoperative detection of PALN after NAT with FOLFIRINOX for pancreatic head adenocarcinoma and its impact on survival, as our policy was to not resect the tumor in such situations. From January 2014 to December 2020, 136 patients with non-metastatic cancer who received neoadjuvant FOLFIRINOX and underwent explorative laparotomy were included. Intraoperative positive PALN were observed in 7 patients (5%). Patients had resectable (n = 5) or locally advanced (n = 2) disease at the time of surgery, but none of them underwent surgical resection. Positive PALN were significantly associated with a lower median number of FOLFIRINOX cycles (4 vs. 6, P = 0.05). There was no significant difference in overall survival between patients with positive loco-regional lymph nodes after resection and patients with non-resection owing to positive PALN (22 versus 16 months, P = 0.16), Overall survival with positive PALN, carcinomatosis, and liver metastasis was 16, 14, and 10 months, respectively (P > 0.05). Our results suggest that NAT may lower PALN involvement. We have modified our policy, positive PALN after NAT are no longer a contraindication to resection, rather a holistic picture of the disease guides management.

Intrahepatic cholangiocarcinoma is a rare disease with a poor prognosis. In patients where surgical resection is possible, outcome is influenced by perioperative morbidity and lymph node status. Laparoscopic liver resection is associated with improved clinical and oncological outcomes in primary and metastatic liver cancer compared with open liver resection, but evidence on intrahepatic cholangiocarcinoma is still insufficient. The primary aim of this study was to compare overall survival for a large series of patients treated for intrahepatic cholangiocarcinoma by open or laparoscopic approach. Secondary objectives were to compare disease-free survival, predictors of death, and recurrence.Patients treated with laparoscopic or open liver resection for intrahepatic cholangiocarcinoma from 2000 to 2018 from 3 large international databases were analyzed retrospectively. Each patient in the laparoscopic resection group (case) was matched with 1 open resection control (1:1 ratio), through a propensity score calculated on clinically relevant preoperative covariates. Overall and disease-free survival were compared between the matched groups. Predictors of mortality and recurrence were analyzed with Cox regression, and the Textbook Outcomes were described.During the study period, 855 patients met the inclusion criteria (open liver resection = 709, 82.9%; laparoscopic liver resection = 146, 17.1%). Two groups of 89 patients each were analyzed after propensity score matching, with no significant difference regarding pre- and postoperative variables. Overall survival at 1, 3, and 5 years was 92%, 75%, and 63% in the laparoscopic liver resection group versus 92%, 58%, and 49% in the open liver resection group (P = .0043). Adjusted Cox regression revealed severe postoperative complications (hazard ratio: 10.5, 95% confidence interval [1.01-109] P = .049) and steatosis (hazard ratio: 13.8, 95% confidence interval [1.23-154] P = .033) as predictors of death, and transfusion (hazard ratio: 19.2, 95% confidence interval [4.04-91.4] P.001) and severe postoperative complications (hazard ratio: 4.07, 95% confidence interval [1.15-14.4] P = .030) as predictors of recurrence.The survival advantage of laparoscopic liver resection over open liver resection for intrahepatic cholangiocarcinoma is equivocal, given historical bias and missing data.

To describe and study the utility of vascular suture-mediated closure systems for large bore arterial access during challenging implantation of liver intra-arterial catheters taking as a reference the conventional procedure involving patients without challenging anatomy. Between January 2017 and January 2019, 61 consecutive patients underwent 65 intra-arterial catheter IAC implantations for colorectal cancer. Twenty-three procedures (35%) considered by the operators with challenging coeliac trunk angulations were treated using a vascular suture technique where a 6-F introducer was used, the other patients were treated with a conventional 4F access technique. Clinical and radiological characteristics of patients, technical success (implantation of catheters allowing safe infusion of chemotherapy) and complications (Common Terminology Criteria for Adverse Events, CTCAE 5.0) were recorded. Mean coeliac trunk angulations were 36.3° (± 14.3) for the vascular closure group and 49.6° (± 17.1) for the conventional group. Technical success of the procedures was 100% for the vascular closure group and 80% in the conventional group (p

Introduction Lors d’une pancreatectomie avec resection veineuse tronculaire, une reconstruction par interposition d’un greffon prothetique en polytetrafluoroethylene (PTFE) peut etre une option meme si un taux de thrombose eleve est souvent craint. Nous rapportons les resultats d’une serie monocentrique de reconstruction par PTFE avec un protocole d’anticoagulation postoperatoire standardise. Methode De 2014 a 2019, 19 reconstruction veineuse par PTFE ont ete realise. La permeabilite prothetique etait evaluee par scanner avant la sortie puis lors des consultations iteratives. Resultats La duodenopancreatectomie cephalique (DPC) representait l’intervention la plus realisee (15 patients, 79 %) et l’adenocarcinome etait l’histologie principale (17 patients, 89 %). La prothese avait un diametre et une longueur mediane de 1 cm et 8 cm, respectivement. La duree mediane de clampage etait de 25 min. La morbidite severe et la mortalite a 90 jours etaient de 21 % et 10 %, respectivement. Aucune infection de prothese n’a ete diagnostique meme en cas de fistule pancreatique ou biliaire. Il n’y a pas eu d’accident d’anticoagulation. Une thrombose precoce ( Conclusion La reconstruction veineuse par une prothese en PTFE au cours d’une pancreatectomie est efficiente et sans risque infectieux. Notre protocole standardise permettait une permeabilite a long terme acceptable. En cas de pancreatectomie gauche, une reconstruction autologue est a recommander.

Background This study aimed to investigate the short- and long-terms outcomes of patients undergoing major hepatectomy (MH) with inferior vena cava (IVC) resection for intrahepatic cholangiocarcinoma (ICC). Methods Data from all patients who underwent MH for ICC with or without IVC resection between 2010 and 2018 were analysed retrospectively. Postoperative outcomes, overall survival (OS), and recurrence-free survival (RFS) were compared in the whole population. A propensity score matching (PSM) analysis and an inverse probability weighting analysis (IPW) were performed to assess the influence of IVC resection on short- and long-terms outcomes. Results Among the 78 patients who underwent MH, 20 had IVC resection (IVC patients). Overall, the mortality and severe complication rate were 8% and 20%, respectively. IVC patients required more extended hepatectomies (p = 0.001) and had increased rates of transfusions (p = 0.001), however they did not experience increased postoperative morbidity, even after PSM. The 1-, 3- and 5-years OS and DFS were 78%, 45%, and 32% and 48%, 20%, and 16%, respectively. IVC was not associated with decreased OS (p = 0.52) and/or RFS (p = 0.85), even after IPW. Conclusion MH with IVC resection for ICC seems to provide acceptable short- and long-term results in a selected population of patients.

Background Sacrificing a replaced right hepatic artery (rRHA) from the superior mesenteric artery is occasionally necessary to obtain an R0 resection after pancreaticoduodenectomy (PD) for pancreatic adenocarcinoma (PA). Preoperative embolization (PEA) of the rRHA has been proposed to avoid the onset of postoperative biliary and ischemic liver complications. Methods Eighteen patients with cephalic PA with an rRHA underwent PEA of the rRHA from 2013 to 2019. The monitoring after embolization and PD was systematic and included a clinical-biological evaluation and a computed tomography scan. This study aimed to determine the feasibility of PEA of the rRHA, postoperative morbidity at 90 days, and quality of oncologic resection after PD. Results Feasibility of PEA was 100% without complications. A PD was performed in 16/18 patients. Mortality was 2/16 with one death after septic shock with hepatic ischemia without an arterial obstruction. Overall morbidity was 44% including one hepatic abscess after hepatic ischemia (6%). Two resections were R1 ( Conclusion PEA of the rRHA before PD was safe and reproducible. PEA of the rRHA followed by en bloc PD resection seems to limit the risk of bilio-hepatic ischemia and could facilitate oncologic resection.

Backgrounds/Aims Surgical resection remains the gold standard in the treatment of colorectal liver metastasis. However, when a patient presents with a deep solitary colorectal liver metastasis (S-CLM), the balance between the hepatic volume sacrificed and the S-CLM volume is sometimes clearly unappropriated. Thus, alternatives to surgery, such as operative and percutaneous radiofrequency ablation (RFA) and microwave ablation (MWA), have been developed. This study aimed to identify the prognostic factors affecting survival of patients with S-CLM who undergo curative-intent liver resection or local destruction (RFA or MWA). Methods We retrospectively identified 211 patients with synchronous or metachronous S-CLM who underwent either surgical resection (n=182) or local destruction (RFA or MWA; n=29) according to the S-CLM size, location, and surrounding Glissonian structures. Results Patients who underwent RFA or MWA had S-CLM of a smaller size than those who underwent resection (mean 19.7 vs. 37.3 mm, p

Introduction La totalisation pancreatique (TP) en urgence apres duodenopancreatectomie cephalique (DPC) pour fistule grade C est une procedure difficile notamment pour les jeunes chirurgiens. Nous rapportons les resultats d’une technique standardisee en 4 etapes etablie afin de permettre une reintervention rapide meme par un chirurgien non-senior. Methode Lorsqu’une reintervention etait decidee, elle etait toujours validee par un chirurgien senior ; le chirurgien non-senior pouvait debuter l’intervention sans attendre l’arrivee du chirurgien responsable, qui venait toujours en renfort. La premiere etape est de deconnecter la gastro-entero-anastomose pour acceder rapidement a l’anastomose pancreatique et permettre ainsi une exposition adequate, particulierement en cas d’hemorragie active. La deuxieme etape, en cas d’anastomose pancreatico-jejunale, est de deconnecter l’anse pancreatique depuis l’anastomose hepatico-jejunale. Troisiemement, la totalisation pancreatique est effectuee avec ou sans preservation des vaisseaux spleniques et de la rate en fonction des conditions locales. Enfin, la quatrieme etape est la reconstruction digestive a l’aide d’une anse en Y, associee a un drainage large. Resultats De 2012 a 2019, 450 patients ont eu une DPC, et une reintervention pour fistule grade C etait decidee chez 30 patients : une TP a ete effectuee chez 21 patients (4,7 %) apres un delais moyen de 12 jours. Les pertes sanguines et la duree operatoire etaient de 600 mL et 4 h respectivement, en moyenne. La mortalite postoperatoire etait de 24 %. Conclusion La mortalite etait elevee mais inferieure a celle rapportee par d’autres series. Notre procedure standardisee semble donc fiable et reproductible, et peut etre particulierement utile pour les chirurgiens non-seniors.

Introduction En 2013, l’anastomose pancreatico-gastrique intussusceptee (PGI) etait decrite comme diminuant le taux de fistule pancreatique mais chez des patients ayant un risque variable. Le but de notre etude etait de comparer l’incidence de la fistule pancreatique cliniquement significative (FPC ; grade B ou C) chez les patients ayant une anastomose a haut risque (updated alternative-fistula risk score [ua-FRS] > 20 %) selon qu’ils aient eu une PGI ou une pancreatico-jejunale (PJ). Methodes De 2013 a 2019, 198 patients consecutifs ayant un ua-FRS moyen de 33 % ont ete inclus. Notre equipe avait une experience avec les PG directe mais une PGI n’etait realisee (n = 33) que lorsque les conditions etaient optimales (pancreas facilement mobilisable et exposition facile) en suivant la technique publiee. Les patients ont ensuite ete apparies sur l’IMC, le diametre du canal pancreatique, et l’ua-FRS avec 165 PJ. Resultats Pour l’ensemble des patients, une FPC etait diagnostiquee chez 42 patients (21 %), et une complication hemorragique chez 30 patients (15 %). La mortalite postoperatoire etait de 4 %. Le taux de FPC dans les groupes PJ et PGI etait de 19 % versus 33 % (p = 0,062), respectivement. Dans le groupe PGI, on notait une incidence plus elevee des FP grade C (24 versus 10 % ; p = 0,036), une duree operatoire plus longue (p = 0,019), et un taux de transfusion plus eleve (p Conclusion La PGI n’etait pas superieure a la PJ pour la prevention des FPC et, dans cette indication, a ete abandonnee par notre equipe.

Resume But Evaluer la presence et l’impact de la mutation K-ras dans une marge de resection histologiquement negative (R0) d’une piece de duodenopancreatectomie cephalique (DPC) pour adenocarcinome resecable de la tete du pancreas. Methode De 2007 a 2010, 22 patients ont eu une DPC en marge saine (R0) pour un ADK resecable d’emblee. Tous les specimens etaient encres et le lit de la veine porte (VP) etait repere par de l’encre bleu ; un echantillon de 2 mm3, incluant de l’encre bleue, etait preleve dans une zone ne comportant pas de tumeur apres verification microscopique. L’ADN de cet echantillon etait extrait et la mutation du gene K-ras etait recherchee : 12 echantillons (55 % ; groupe kras+) presentaient une mutation K-ras dans la marge veineuse et 10 echantillons (45 % ; groupe kras−) n’avaient pas de mutation K-ras dans cette meme marge. Resultats Les 2 groupes etaient comparables. Les survies globales a 3 ans des patients des groupes kras+ versus kras− etaient de 0 % versus 17 % (p = 0,03), respectivement. Les medianes de survies des patients des groupes kras+ versus kras− etaient de 16 mois versus 25 mois (p = 0,04 ; intervalle de confiance a 95 % [1,11–1,88]), respectivement. Conclusion La mutation K-ras dans une marge histologiquement negative (R0) etait presente dans plus de la moitie des specimens avec un impact negatif sur la survie.

BACKGROUND The occurrence of post-operative pancreatic fistula (POPF) after pancreaticoduodenectomy is a challenging issue. The aim was to identify variables on preoperative computed tomography (CT) scan, useful to predict clinically significant POPF (grades B-C) after pancreaticoduodenectomy. METHODS Patients presented POPF after pancreaticoduodenectomy were included from two tertiary referral centres. B/W ratio was defined by ratio of pancreas body thickness (B) to main pancreatic duct (W). The predictive parameters of POPF on CT scan were assessed with a receiving operator characteristics (ROC) curve and intrinsic characteristics. RESULTS Between 2010 and 2013, 186 patients who underwent pancreaticoduodenectomy were included. POPF occurred in 25% of them, and was clinically significant in 13%. After univariate analysis, endocrine tumours (P = 0.03), main pancreatic duct size (P 3.8 increased the rates of post-operative haemorrhage (odds ratio = 4.3 (1.4-13.2), P = 0.01), and reintervention (odds ratio = 3.4 (1.2-9.6), P = 0.02). CONCLUSIONS B/W ratio superior to 3.8 assessed on preoperative CT scan may be an easy tool to predict clinically significant POPF after pancreaticoduodenectomy.

Background Transabdominal laparoscopic proctectomy (LAP) for rectal cancer was associated with postoperative recovery improvement. Early studies showed favorable short-term results of endoscopic transanal proctectomy (ETAP), with low conversion rates to open procedures. We aimed to compare efficacy, morbidity, and functional outcomes of ETAP to standard LAP for low rectal cancer. Study Design From 2008 to 2013, 72 consecutive patients received proctectomy and coloanal manual anastomosis for low rectal adenocarcinoma. Thirty-four patients had transanal endoscopic proctectomy, and 38 patients underwent the standard laparoscopic procedure. Results When compared with the LAP group, the ETAP group demonstrated a lower conversion rate to open procedures (23.7% vs 2.9%, respectively; p = 0.015), shorter in-hospital stays (9 vs 8 days, respectively; p = 0.04), and a lower readmission rate (13.2% vs 0%; p = 0.03). Overall postoperative morbidity rates for the LAP and the ETAP groups (36.8% vs 32.4%, respectively; p = 0.69) and functional results (Kirwan score 1/2, 73.7% vs 73.5%, respectively; p = 0.85) were comparable; additionally, we found similar oncologic quality criteria (R1 resection 10.5% vs 5.9%, respectively; p = 0.68; grade 3 mesorectal integrity 52.6% vs 55.9%, respectively; p = 0.66). Disease-free survival of 24 months (Kaplan-Meier estimation) was comparable in the 2 groups: 86% in the ETAP group vs 88% in the LAP group; p = 0.91. At the date of last follow-up, 91.2% of ETAP patients and 92.1% of LAP patients were free of stoma. Conclusions The endoscopic transanal approach could facilitate mesorectal excision and improve short-term outcomes without impairing the oncologic quality of the resection or mid-term functional and oncologic results.

The decision to operate on hematology patients with abdominal emergencies can be difficult, as neutropenia and thrombocytopenia are common and the usual causes of abdominal pain are broad. We conducted a retrospective observational study including all hematology patients undergoing emergency abdominal surgery between January 1998 and January 2013. Of the fifty-eight consecutive patients included in the study, nineteen (33%) underwent an operation during the neutropenia period. In the multivariate analysis, a laparotomy after 2002 was protective (HR: 0.05; 95%CI: 0.001-0.24), whereas preoperative septic shock (HR: 8.58; 95%CI: 2.25-32.63) and use of dialysis (HR: 6.67; 95%CI: 2.11-21.07) were independently associated with hospital mortality. Surgery during neutropenia or thrombocytopenia was not associated with prognosis. In hematology patients, emergency abdominal surgery is associated with encouraging hospital survival rates. Surgery should be performed prior to septic shock, regardless of whether neutropenia or thrombocytopenia is present.

But Evaluer le devenir des patients avec un adenocarcinome du pancreas localement avance (ADKLA) non reseque en fonction de la strategie therapeutique adoptee. Materiel Entre 2010 et 2017, 234 patients ont eu une chimiotherapie d’induction (FOLFIRINOX pour 72 % d’entre eux) pour ADKLA et n’etaient pas accessibles a la resection chirurgicale ; 47 patients etaient metastatiques a la reevaluation et chez 187 il etait decide soit la poursuite de la chimiotherapie (61 patients, groupe CT), soit une radiochimiotherapie (126 patients, groupe RCT) de cloture. Resultats Les deux groupes etaient comparables selon les caracteristiques cliniques. Une progression metastatique etait observee chez 90 patients (71 %) du groupe RCT et 26 patients (43 %) du groupe CT (p = 0,01). Les survies globales des 2 groupes etaient comparables (survie mediane de 19 mois) mais la survie sans progression etait meilleure pour le groupe RCT (mediane 13,3 vs 9,6 mois, p Conclusion Chez les patients avec un ADKLA non resecable et non metastatique apres chimiotherapie d’induction, la poursuite d’une CT ou la realisation d’une RCT procuraient une survie globale identique. La RCT entrainait une survie sans progression plus longue avec moins de toxicite.

Background The current study aimed to evaluate the outcomes of patients with unresectable non-metastatic locally advanced pancreatic adenocarcinoma (LAPA) who did not benefit from resection considering the treatment strategy in the clinical settings. Methods Between 2010 and 2017, a total of 234 patients underwent induction chemotherapy for LAPA that could not be treated with surgery. After oncologic restaging, continuous chemotherapy or chemoradiation (CRT) was decided for patients without metastatic disease. The Kaplan–Meier method was used to determine overall survival (OS), and the Wilcoxon test to compare survival curves. Multivariate analysis was performed using the stepwise logistic regression method. Results FOLFIRINOX was the most common induction regimen (168 patients, 72%), with a median of 6 chemotherapy cycles and resulted in higher OS, compared to gemcitabine (19 vs. 16 months, hazard ratio (HR) = 1.2, 95% confidence interval: 0.86–1.6, P = .03). However, no difference was observed after adjusting for age (≤75 years) and performance status score (0–1). At restaging, 187 patients (80%) had non-metastatic disease: CRT was administered to 126 patients (67%) while chemotherapy was continued in 61 (33%). Patients who received CRT had characteristics comparable to those who continued with chemotherapy, with similar OS. They also had longer progression-free survival (median 13.3 vs. 9.6 months, HR = 1.38, 95% confidence interval: 1–1.9, P Conclusions The median survival of patients who could not undergo surgery was 19 months. Hence, CRT should not be eliminated as a treatment option and may be useful as a part of optimised sequential chemotherapy for both local and metastatic disease.

Objective To determine the effect of clinical status (weight variation and performance status [PS]) at diagnosis and during induction treatment on resectability and overall survival (OS) rates in patients with borderline resectable (BRPC) or locally advanced pancreatic cancer (LAPC). Methods From 2005 to 2017, 454 consecutive patients were diagnosed with LAPC or BRPC. We evaluated the PS (0–1 or 2–3), body mass index at diagnosis, and weight loss (WL) > 5% at initial staging and after induction treatment and separated continuous weight loss (CWL) from weight stabilization. Results A total of 294 patients (64.8%) presented with WL, and 57 patients (12.6%) presented with a PS of 2–3. At restaging, 60 patients (13.2%) presented with CWL. Independent factors that poorly influenced the OS were a PS of 2–3 at diagnosis (P < .01), CWL at restaging (P < .01), and absence of resection (P < .01). Factors independently impeding resection were LAPC (P < .01), PS > 1 at diagnosis (P < .01), and CWL (P = .01). In total, 142 patients (31.3%) underwent pancreatectomy. Independent factors that poorly influenced the OS in the resected group were PS > 0 at diagnosis (P = .01) and obesity (P < .01). For the 312 unresected cancer patients (68.7%), CWL (P < .01) was identified as an independent factor that poorly influenced the OS. Conclusion Clinical parameters that are easy to measure and monitor are independent factors of poor prognosis. The variation of weight during the induction treatment, more than WL at diagnosis, significantly precluded resection and was an independent factor of shorter OS in unresected patients.

A new neoadjuvant regimen, together with more aggressive surgeries, appears to have increased the resectability rate in patients with pancreatic ductal adenocarcinoma (PDAC). Our study aimed to evaluate the outcomes of patients who underwent venous resection (VR) during pancreatectomies for PDAC.Between 2005 and 2017, 130 patients underwent pancreatectomies with type 3 or 4 (i.e., segmental resection without or with graft interposition, respectively) VR for PDAC. Patients' characteristics, surgical techniques, perioperative management, pathological findings, and outcomes were recorded and compared during 2 inclusion periods: the landmark year for the introduction of the FOLFIRINOX regimen and the hyperspecialization of our pancreatic-surgery team was 2010.Performance of pancreatectomies with VR steadily increased through the 2 inclusion periods. In the overall series (n = 130), the median overall survival time and the 5-year survival proportion were 26.3 months and 21%, respectively. Upon multivariate analysis, ASA score 3 (P = 0.01) and R1 resection margins (P 0.01) were found to be negative independent factors influencing survival. Patients who underwent upfront VR (n = 47) had survival rates similar to those of patients who received neoadjuvant treatment (n = 83). After 2010, more complex VR were performed; however, no difference was found between the 2 periods with respect to postoperative courses, pathologic findings, or survival after a matching process based on patients' characteristics and tumor stages.Over the last 2 decades, VR during pancreatectomy has been confirmed as a safe procedure despite the increase in technical complexity. Disappointingly, we did not observe any dramatic survival improvement.

Purpose This study proposes and details a simple and inexpensive protective technique of wrapping the hepatic and gastroduodenal artery stumps with a peritoneal patch during pancreatoduodenectomy (PD) in order to decrease post-pancreatectomy hemorrhage (PPH). Methods Among the 85 patients who underwent PD between July 2020 and March 2021, 16 patients with high-risk pancreatic anastomosis received a peritoneal patch. The Updated Alternative Fistula Risk Score (ua-FRS) was calculated. Post-operative pancreatic fistula (POPF) and PPH were diagnosed and graded according to the International Study Group of Pancreatic Surgery. The mortality rate was calculated up to 90 days after PD. Results The mean ua-FRS of the 16 patients was 43% (range: 21-63%). Among them, 6 (38%) experienced clinically relevant-POPF, and a PPH was observed in two patients (13%). In these two patients who required re-intervention, the peritoneal patch was remarkably intact, and neither the gastroduodenal stump nor hepatic artery was involved. None of the patients experienced 90-day mortality. Conclusion Although the outcomes are encouraging, the evaluation of a larger series to assess the effectiveness of the peritoneal protective patch for arteries in a high-risk pancreatic anastomosis is ongoing.

Objectif Evaluer les suites operatoires et la survie de patients operes d’hepatectomie majeure (HM) avec resection de la veine cave inferieure (VC) pour cholangiocarcinome intrahepatique (CIH). Materiel et methode Entre 2010 et 2018, 18 patients ont eu une HM avec resection de la VC (groupe VC) et ont ete compares aux patients operes d’HM pour CIH (n = 60) durant la meme periode. Les resultats postoperatoires, la survie globale (SG) et sans recidive (SSR) ont ete analysees. Resultats Douze patients ont eu une resection laterale et 6 patients un remplacement cave prothetique. Les suites operatoires ainsi que la mortalite a 90 jours etaient comparables. Seule la transfusion postoperatoire etait identifiee comme facteur de risque de mortalite a 90 jours (p = 0,015). La mediane de survie dans le groupe VC et controle n’etait pas differente statistiquement (48 mois vs 27 mois p = 0,32) de meme que le taux de SG a 1-, 3- et 5 ans (85/81 %, 50/23 %, et 30/8 %). La mediane de SSR etait de 11 mois dans les 2 groupes. L’atteinte ganglionnaire etait le seul facteur de risque de diminution de la survie globale (p = 0,042). Conclusion La resection de la VC au cours d’une HM pour CIH ne pejore pas les resultats a court et long terme.

But Evaluer les marges de resection chez les patients ayant une duodenopancreatectomie cephalique pour cholangiocarcinome (CC) ou adenocarcinome pancreatique (ADK), en utilisant un protocole d’analyse standardise. Materiel De 2010 a 2018, 355 specimens encres d’ADK (n = 288) ou de CC (n = 67) ont ete analyses. Une resection R1 etait definie par un seuil de 0 mm, 1 mm, ou 1,5 mm. Resultats Les patients avec un CC etaient plus souvent des hommes (p = 0,028), plus âges (p = 0,033), necessitaient plus frequemment un drainage biliaire (p Conclusion Avec un protocole d’analyse standardisee, le taux de resection R1 etait comparable chez les patients avec ADK et CC.

But Evaluer les criteres predictifs d’une survie postoperatoire inferieure a 1 an chez les patients ayant adenocarcinome du pancreas localement avance necessitant une resection vasculaire, apres chimiotherapie d’induction par Folfirinox. Materiel Entre 2011 et 2018, 130 patients ont eu une pancreatectomie apres Folfirinox d’induction ; 75 necessitaient une resection vasculaire (veineuse tronculaire 93 %, avec resection arterielle concomitante 15 % ou isolee 7 %) et composaient notre groupe d’etude. Resultats La mortalite postoperatoire etait de 8 % a 90 jours. Les facteurs associes a une survie postoperatoire inferieure a 1 an en analyse univariee etaient : une perte de poids > 5 % (p = 0,05), un PS ≥ 2 (p = 0,02), un CA19-9 preoperatoire > 300 (p = 0,012), une hemorragie postoperatoire (p = 0,014), le stade T (p = 0,021), et nombre de ganglions envahis (p = 0,023). En analyse multivariee, la perte de poids (OR = 3,61, [1,13 ; 11,5], p = 0,031), et le CA19-9 > 300 (OR = 5,58, [1,15 ; 27,1], p = 0,034) etaient des facteurs preoperatoires independants d’une survie inferieure a 1an. Conclusion Une perte de poids > 5 % et un CA19-9 > 300 etaient predictif d’une survie postoperatoire inferieure a 1 an. Ces 2 criteres devraient etre integres dans l’arbre decisionnel lorsqu’une pancreatectomie avec reconstruction vasculaire est envisagee.

Background According to knowledge, patients with resectable pancreatic adenocarcinoma (PA) should receive adjuvant gemcitabine-based chemotherapy. Thus, the tumour node metastasis (TNM) classification is not used to determine post-operative treatment but rather only to establish patient prognosis. However, the TNM classification does not include strong factors influencing survival, such as perineural invasion or margin status. This study compared the survival of patients with very similar tumours. Methods From 1997 to 2007, 118 patients underwent pancreatectomy for PA. Twenty-six patients (22%) had long-term survival (>5 years; LTS group). According to the major prognostic factors of PA, we matched (1:1) patients in the LTS group with patients who did not have long-term survival (

Objective To assess the accuracy of pre‐operative staging in patients with peripheral pancreatic cystic neoplasms (pPCNs). Methods From 2005 to 2011, 148 patients underwent a pancreatectomy for pPCNs. The pre‐operative examination methods of computed tomography (CT), magnetic resonance imaging (MRI), endoscopic ultrasonography (EUS) were compared for their ability to predict the suggested diagnosis accurately, and the definitive diagnosis was affirmed by pathological examination. Results A mural nodule was detected in 34 patients (23%): only 1 patient (3%) had an invasive pPCN at the final histological examination. A biopsy was performed in 79 patients (53%) during EUS: in 55 patients (70%), the biopsy could not conclude a diagnosis; the biopsy provided the correct and wrong diagnosis in 19 patients (24%) and 5 patients (6%), respectively. A correct diagnosis was affirmed by CT, EUS and pancreatic MRI in 60 (41%), 103 (74%) and 80 (86%) patients (when comparing EUS and MRI; P = 0.03), respectively. The positive predictive values (PPVs) of CT, EUS and MRI were 70%, 75% and 87%, respectively. Conclusions Pancreatic MRI appears to be the most appropriate examination to diagnose pPCNs accurately. EUS alone had a poor PPV. Mural nodules in a PCN should not be considered an indisputable sign of pPCN invasiveness.

Predictors of survival following pancreaticoduodenectomy (PD) are well described for pancreatic cancers but are less detailed in ampullary (AC), bile duct (BDC) and duodenal cancers (DC). We therefore sought to evaluate the long-term results of PD for AC, BDC and DC, and to determine for each tumour the predictive factors of survival. Medical charts of patients operated on between 2001 and 2011 were retrospectively reviewed. Univariate and multivariate analyses were performed to determine predictors of survival. One hundred thirty-five patients were identified. Mean follow-up was 47 ± 33 months. Median survival was not reached for DC and was 66 and 24 months for AC and BDC, respectively. Two-year and five-year survival rates were 80 and 51 % for DC and 69 and 51 % for AC, respectively. BDC had a significantly poorer prognosis, with two-year and five-year survival rates of 51 and 34 %, respectively. Predictors of survival were weight loss, N stage and International Union Against Cancer (UICC) stage for AC, T stage and resection margin status for BDC and N stage for DC. AC, BDC and DC display distinctive predictors of survival related to the biological aggressiveness. Preoperative malnutrition worsens the prognosis. The effect of adapted nutritional management on the survival improvement has to be studied.