Postoperative pancreatic fistula (POPF) is the main driver of surgical morbidity after pancreatoduodenectomy (PD). The aim of the present study is to compare total pancretectomy (TP) and primary pancreatic anastomosis (PA) in a cohort of extremely high-risk patients, with regards to postoperative outcomes and quality of life (QoL).

The study will be organized as a multicentric randomized controlled trial. When a patient will be considered eligible preoperatively, he or she will be enrolled in the present study. The patient's risk will be allocated intraoperatively after the demolitive time and before pancreatic anastomosis, according to the following criteria:

• Major Criteria: main pancreatic duct diameter ≤3mm; soft pancreas
• Minor Criteria: bleeding stump; friable stump; posterior/eccentric duct; invisible duct; deep pancreas; intraoperative acute pancreatitis; FRS 9-10
• Interobserver agreement ≥2 Surgeons

Patients presenting with all major criteria (stage D according to Schuh et al.) and at least one among the minor criteria will be considered eligible. At least two surgeons must confirm eligibility, according to interobserver agreement regarding the above-mentioned criteria and ethical applicability of randomization (ethical "guarantor"). Eligible patients will be randomized to receive PA or TP.

Pre-operative, intra-operative and post-operative data will be recorded prospectively by the P.I. and by his collaborators.

Preoperative care Preoperative care will follow institutional standards, according to each center involved. After obtaining the informed consent for elective pancreatic resection and after adequate counselling, the study will be proposed to the patient. If the patient will accept, the informed consent will be obtained. The physician who will get the informed consent for the procedure will also be responsible for the consent to the study. Randomization lists will be provided for each Center and for each randomized group. All patients will be preoperatively checked for diabetes by measuring fasting blood glucose (FBG) and glycated hemoglobin (Hb1Ac), C- peptide, and for pancreatic exocrine insufficiency by associated clinical criteria and fecal elastase-1 (FE-1) determination.

Intraoperative setting The resection phase will be carried out according to clinical practice at each participating center. At the time of pancreatic anastomotic reconstruction, the risk will be calculated on the basis of the presence of major and minor criteria The operating surgeon will decide whether the patient can be included in the study or must be excluded for the presence of any drop-out criteria.

The interobserver agreement of at least another surgeon will serve as a confirmation for the inclusion in the trial (ethical "guarantor"). According to the presence of two major and one or more minor criteria patients will be included in the study and randomized in two groups. A picture of the pancreatic remnant will be taken intraoperatively before anastomosis/totalization. Both patients undergoing open and minimally invasive surgery can be considered eligible.

PA Group PA will be carried out according to the techniques adopted by the participating Centre, either pancreatico-jejunostomy (PJ) (i.e. dunking PJ, Cattel-Warren duct-to-mucosa PJ, Blumgart PJ) or pancreatico-gastrostomy (PG) will be considered eligible. Any mitigation strategy (i.e. externalized transanastomotic stent, use of glues/biological matrices to protect the anastomosis, surgical feeding jejunostomy, prophylactic hydrocortisone/somatostatin administration) can be used according to the Center practice. The other two anastomosis, hepaticojejunostomy and duodenojejunostomy (in case of Longmire-Traverso PD) or gastrojejunostomy (in case of Kausch-Whipple PD), will be carried out as usual according to each Institution's operative standards.

At least one surgical drain will be placed in the retroperitoneum in all patients.

TP Group TP will be carried out according to each Institution's operative standards. Preservation of the spleen will be considered whenever possible according to Kimura technique. Either ligation or preservation of gastric vessels (right/left gastric artery/vein) will be allowed according to clinical necessity but will be recorded and correlated with postoperative outcomes.

The reconstruction phase will be carried out according to each Institution's operative standards. One or more surgical drains can be left in place according to surgeon's preference.

Postoperative care After the procedure, the patient will be admitted to the ICU or in the ward. The management of intravenous fluids, nasogastric tube, bladder catheter and postoperative analgesia will take place as usual according to each Institution's standards of care. In the PJ Group, the amylase value of drain fluids will be checked on postoperative day (POD) 1, 3 and at any POD if it will help to diagnose a still undiscovered POPF. Surgical drain will be managed according to clinical judgment and each Institution's clinical standards. During hospitalization, all patients will receive specialistic evaluation to assess and possibly treat the occurrence of new onset diabetes or the worsening of pre-existing diabetes. Pancreatic exocrine insufficiency will be treated with oral supplementation of pancreatic enzymes if needed. Patients in the TP Group will be vaccinated 1 month after complete functional recovery against Pneumococcus, Hemophilus influenzae group B, and Meningococcus to minimize the likelihood of developing post-splenectomy sepsis. All patients will receive an outpatient follow-up, 1 month after discharge and every 6 months for 2 years. During follow-up, glycemic control, nutritional status, and possible symptoms of exocrine insufficiency will be assessed, and patients will receive specialistic assistance if needed. QoL will be registered using specific questionnaires. An oncologic evaluation will determine the indication for administration of adjuvant therapy, in case of malignancy.

All patients who will be alive after at least 12 months of follow-up will be enrolled in the cross-sectional study of quality of life. All the eligible patients who are not able to attend outpatient visits will be contacted by telephone before receiving the 5 questionnaires by mail. Patients who will not respond within 1 month will be contacted again by telephone.

SAMPLE-SIZE CALCULATION:

A recent metanalysis by Schuh et al. reported a POPF rate around 23% in stage D patients (main pancreatic duct diameter ≤3mm, soft pancreas) but no data were available regarding major morbidity12. Patients that will be considered eligible for the present trial will have further features increasing their risk of POPF and related major morbidity, compared to stage D patients. Based on the current literature and on a recent retrospective study by the Verona group, considering only a cohort of patients with extremely high FRS (FRS 9-10), the rate of major morbidity can be estimated around 55% after PD and 19% after TP.11 The original risk score proposed in this trial is extrapolated based on previous literature and expert opinions. For this reason, there are no data directly available for estimation of major morbidity in this rare subset of patients, for which we estimate a rate of Clavien-Dindo ≥3 of around 40%.

Considering a 1:1 allocation between the groups, a sample size of 49 patients per group would allow a two-sides, two-sample test for binomial proportions to detect a difference in Clavien-Dindo ≥3 of 25% (40% vs 15%) with 80% power (1-β) and an error α of 0,05. The study has a group sequential design allowing for interim analyses at pre-specified timepoints with possible early stopping for efficacy or futility in case of an overwhelming large or small effect, respectively. Considering that extremely high-risk cases represent around 7% of the total amount of PDs performed and that the fistula risk can only be assessed intraoperatively, we expect to approach preoperatively around 1300 patients.

DURATION:

Since the Department of General and Pancreatic Surgery performs about 250 PD per year and the study will be proposed to at least other 2 international referral centers, the primary endpoint will be reached after 25 months: 22 months of patients' enrollment and 3 months of follow-up. Long term analysis of QoL and pancreatic insufficiency will require additional 24 month of follow-up. Time for data analysis must be considered negligible. All the patients undergoing PD will be enrolled for the study if inclusion and exclusion criteria will be respected. The study will be discontinued in case of reaching the statistical significance of the primary endpoint, or in case of suspension by the coordinators or by the authorities.

Inclusion Criteria:

• Patients older than 18 years
• All patients scheduled for PD for all kind of pancreatic diseases
• Patients able to give their informed consent
• Patients undergoing PD (Kausch-Whipple or Longmire-Traverso)
• Patients presenting two major and at least one minor criteria (Major criteria: Main pancreatic duct diameter ≤3mm; Soft pancreas. Minor criteria: Bleeding stump; Friable stump; Posterior/Eccentric duct; Invisible duct; Deep pancreas; Intraoperative acute pancreatitis; FRS 9-10)
• Two or more surgeons confirming eligibility
• PA or TP with or without spleen preservation (Kimura technique). These techniques are consistent with clinical practice; any other procedure will be a deviation from the protocol

Exclusion Criteria:

• Informed consent withdrawal
• Impossibility to undergo surgery for any reason
• Main pancreatic duct of the pancreatic neck/body >3mm at preoperative imaging (CT scan or MRI)
• PD not performed for any reason
• Absence of two major criteria
• Absence of at least one minor criteria
• Absence of interobserver agreement between at least 2 surgeons
• More than 1 extension of resection to pancreatic neck due to pancreatic margin positivity
• Wrong randomization