Early POstoperative BIomarkers in PancreatoDuodenectomy: a Spanish Nationwide Study

Overview

AIMS: To recruit a large nationwide Spanish series in order to register PF rate and other common morbidity after PD, and to perform external validation of the aforementioned score, as well as to analyze other postoperative blood parameters and biomarkers associated with complications. METHODS: Observational prospective and multicentric cohort study to perform external validation of the above-mentioned score. All Spanish centers/units performing PD are invited to include participants. Patients will be consecutively recruited during an 8-10 months period, regardless of their annual volume of pancreatic surgery. Study variables will be hemogram parameters on POD1 and POD2 (specifically lymphocytes), other parameters and biomarkers (RCP, lactate, procalcitonin, amylase, lipase, albumin) and the common variables concerning PD studies. LIMITATIONS: Heterogeneity in perioperative management and in blood analysis measuring since this is a multicenter study. Possibility of underestimating the PF rate in patients without surgical drainage. Finally, the cases of mini-invasive approach or pancreatogastrostomy will receive a specific subgroup analysis since the score was designed on a series of open PD and pancreatojejunostomy.

BACKGROUND: Complications after pancreatoduodenectomy (PD) are common and pancreatic fistula (PF) is among the most relevant ones. There are few studies analyzing blood parameters determined in the early postoperative period which behave as morbidity predictors. One of them defined a new and simple prognostic score based on lymphocytes count on the first postoperative day (POD1) and reactive C protein (RCP) on POD2. The internal validation statistics of the score designed in the mentioned work were satisfactory. However, that score was described in a short unicentric series EXPECTED SAMPLE SIZE: Expected sample size is approximately 500 patients, which will show error \<7.2% (with 95% confidence) for estimation score´s sensitivity and specificity. We have assumed that the PF rate is around 30%. QUALITY CONTROL: The STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines for the correct communication and publication of observational studies will be followed. Likewise, in order to validate the aforementioned score, Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) will be added. The only expected failure of the 20 items that compose the TRIPOD is 6b ("notify the actions for masking the analysis"), since it is not applicable in our case. The inclusion of patients operated on with a mini-invasive approach (laparoscopic or robotic) or the use of other anastomoses (pancreatogastrostomy) or reconstruction (Roux-en-Y) differs from the technique used in the series with which the score was originally defined (allof them, open PD with pancreato-jejunostomy and, mostly part, tutored ducto-mucosal suture and Child's single loop reconstruction). If discrepancies are detected, the corresponding subgroup analysis will be performed. Cases of conversion from a mini-invasive to open approach will be counted as such (intention-to-treat analysis).