Venous Resection in Patients Undergoing Pancreatic Surgery

Objectives: To assess the perioperative technical, organisational, and treatment-related factors associated with increased risk for early and late venous thrombosis in patients undergoing portal vein/superior mesenteric vein resection (PV/SMV) during pancreatectomy. To assess the perioperative factors associated with lowest thrombosis rate for each type of venous reconstruction, adjusted for the size of the resected segment. Introduction Pancreatic surgery is the only treatment in the multimodality approach to pancreatic cancer sine qua non long- term survival cannot be achieved(1). Pancreatectomy with venous resection (PVR) of the portal vein and the superior mesenteric vein (PV/SMV) to address tumors with venous involvement has become standard of care in pancreatic surgery for the past 20 years(2) (3). Larger series and meta-analyses show that long-term survival beyond the 5-year threshold can be achieved with PVR, but with somewhat inferior survivor compared to standard pancreatic resections without venous resection (4)(5)(1). That is most likely due to the fact that tumors necessitating PVR are generally larger, less differentiated, and more likely to express unfavourable features such as perineural infiltration (3,6-8). Recent publications point out that PVR can be performed in pancreatic centers around the world with comparable morbidity and mortality, irrespective of the countries' income status (9). Also, benchmark outcomes for PVR in low-risk patients have been established, defining that the accepted risk for venous thrombosis should be lower than 14% at discharge (10). Despite that, a questionnaire among expert surgeons reveals that even if PVR is considered a standard procedure, surgeons are still concerned about significant morbidity. Also, the perception as to what proportion of patients require venous resection, how it should be performed and what anti-thrombotic prophylaxis should be given varies widely (11-13). The reported short-term thrombosis rates after different types of venous resections and reconstructions vary from under 2 to over 25% (2,14-16),(17). There is no consensus as to which type of venous reconstruction is associated with most favorable short- and long-term outcomes, what specific surgical manoeuvres and graft materials decrease the overall complication and the vein-resection specific complications rates and which anti-thrombotic prophylaxis is safest (associated with lowest thrombosis rates and bleeding complications). Although some studies suggest that grafts are associated with higher thrombosis rates, there is no profound investigation as to how to avoid the use of grafts or, if necessary, what grafts are associated with least thrombotic risks while not causing increased patient morbidity (18,19). The purpose of this study is to evaluate different technical manoeuvres, anti-thrombotic prophylaxis strategies and organisational features in a large cohort of patients undergoing PVR in median and large-volume centers in order to assess which factors contribute to vein resection-related specific morbidity. Methodology The study design is an international multi-center observational cohort study, with retrospective and prospective part, including retrospectively patients who underwent pancreatic surgery with PV/SMV resection from January 2018 to August 2025 and prospectively from January 2026 to June 2027. Perioperative factors, related to technical decisions, perioperative resuscitation, institutional organisational specificities, and anti-thrombotic prophylaxis, associated with increased risk for early (30-day), intermediate 30 days-6 months and late venous thrombosis (\>6months) will be analysed (20) (21). The thrombosis rate among the four types of venous reconstructions will be compared while adjusting for the length of the resected vein segment. Center eligibility criteria: Median and high-volume pancreatic centers (\>20 pancreatic resected /year). Inclusion criteria: Patients undergoing curative pancreatic resection with simultaneous PV/SMV resection, Any final pathology, Date of pancreatic surgery with PVR between January 1st, 2018 - August 31st, 2025 (retrospectively) Date of pancreatic surgery June 1st, 2026 - June 30th, 2027 (prospectively) Exclusion criteria: Distant metastases (M1) Age \<18 years R2 resection Sample size: The project intends to include a minimum of 1000 patients, but aiming at 2500, with expected venous thrombosis of about 250 patients. Endpoints: PV/SMV thrombosis at 30 days, 31 days-6 months, and \>6 months. Data collection: Institution characteristics: (yearly volume of pancreatic resections, type of hospital - university/community; specialty of surgeon performing resection - pancreatic/vascular/transplant) Patient characteristics (age at surgery, sex, BMI, Charlson comorbidity index, genetic or sporadic pro-thrombotic disease, chronic antiplatelet or anticoagulant therapy). Tumor characteristics at diagnosis and in the preoperative setting (size, location, degree of vessel involvement). Neoadjuvant chemotherapy (yes/no, type, duration). Surgical operative data (Cattell-Braasch manoeuvre, type of procedure (PD, DP, TP), length of surgery, intraoperative blood loss, intraoperative transfusion, preferred method of haemostasis (sealing devices, electrocautery, clips, ligation)). Technical details of vascular resection (resected vessel, length of resected segment (cm), type of vessel reconstruction (I-II-III-IV (type 1 - longitudinal/transversal, type 4 - type of graft), type of suture and eventual extra stiches, clamping time, sacrificed/reimplanted vessel (coronary vein, inferior mesenteric vein, splenic vein, colonic vessels; place of reinsertion), SMA clamping, heparin flush during reconstruction, growth factor yes/no, control of patency after reconstruction - blood flow measurement, palpation, inspection), vessel-specific imaging on POD 1, 2-7 - type, result. Perioperative anti-thrombosis prophylaxis (preoperative/intraoperative/postoperative anti-thrombosis prophylaxis: drug choice (LMWH, heparin, anti-aggregation drug), dose on POD 1-7, length of prophylaxis, fluid balance/postoperative body weight on POD 1 and 2). Post-operative complications (Clavein-Dindo grading, POPF, DGE, PPH (site, timing) according to ISGPS, bile leak, chyle leak, reoperation, medical complications). Venous-resection-related complications (venous thrombosis: time of detection, location, grade of occlusion (partial, subtotal/total), presentation (symptoms), method of detection, treatment (medical (type), reoperation, organ ischemia - treatment, long term patency at 30 days, 6 months; bleeding). Histological features (tumour type, tumor size, nodal involvement, resection margins, V/L/PN 0/1, R0/1, margins of resected vessel. depth of vessel involvement). Survival (DFS, OS, site of recurrence) Definitions: PVR: pancreatectomy with simultaneous PV/SMV resection Venous thrombosis: partial, subtotal or total venous obstruction presented as intraluminal contrast irregularity (radiologically) with/without signs of congestion/ischaemia, or operative finding of occluding thrombus. Venous reconstruction types (according to ISGPS Type 1: partial venous resection with direct closure (tangential or transversal), Type 2: partial venous resection and coverage with patch, (any type) Type 3: segmental resection with direct primary venous anastomosis, Type 4: segmental resection with any type of interposition graft. Policy Securing Confidential center specific data Data specific to particular centers will not be published. However, the centers reserve the rights to publish their own data. Study data coordination No data will be revealed, submitted or published, without the specific grant of authorization from each center. The attributed coordinator from each center will collaborate with the study coordinators (Dr. Luana Genova, and Dr. Tom Van Ravens and Dr. Canan Yasar), to aid the collection of the data. Authorship Authorships will be attributed according to the recommendations from the international committee of medical journal editors (ICMJE). The first three authors will be the coordinators of this project (LG, TVR, CY). The last three authorship positions are reserved for the principal investigators (TH, SM, ER). All other authors from the participating centers will be listed in alphabetical order. A minimum of 20 patients per center is required for authorship. If the institution contributes more than 50 patients to this study, an additional authorship is available. For 100 patients - 3 authorships, for 200+ patients - 4 authorships. For centers including more patients, but not reaching the next tier, one collaboratorship will be added. We ask each participating institution to submit their retrospective data to the lead coordinators of this project (genova.luana@hsr.it, t.w.van\_ravens@lumc.nl , canan.yasar@vgregion.se by May 31st, 2026 Further use of cohort data: Possibly and hopefully, future studies may emerge based on the collected data and centers will be additionally asked for participation. Governance; Data will be collected via a secure, password protected, and encrypted data management system. Data will not be published with hospital identifiers. The data will be merged into a new database and exported as de-identified data for the statistical analysis. Data Collection; Local collaborators: All hospitals may have up to one local investigator; every center investigator will be in regular contact with the designated study coordinators (Luana Genova, Tom Van Ravens, Canan Yasar) and will be responsible for: * Gaining local research ethics approval if required * Identifying and including all eligible patients * Accurately collect baseline and follow-up data * Submit data in the online secure database (RedCap) Prospective data will be continuously collected and uploaded in RedCap. Deadline for participation and data collection: May 31st, 2026.