Reduced Intra-operative Blood Loss in Pancreaticoduodenectomy for Pancreatic or Peri-ampullary Tumors; Monocentric Trial on Standard Open Versus Minimally Invasive Surgery

Pancreaticoduodenectomy (PD) is the only therapeutic option to cure patients suffering from pancreatic head or peri-ampullary tumors. The standard approach in PD is open surgery (OPD). With advancing technology, data from expert centers suggest minimally invasive or laparoscopic PD (LPD) to be safe and feasible, though randomized studies are lacking to show the clinical benefits of LPD vs. OPD. Advantages of minimally invasive surgery over open surgery are reduced tissue damage, surgical trauma and immunosuppression. The general objective of this study is to compare the intra-operative efficacy of LPD vs. OPD, and in particular the amount of intra-operative blood loss. Study design Monocentric randomized trial to compare the amount of intra-operative blood loss in LPD vs. OPD. Patients expected to undergo portal vein resection/reconstruction or any simultaneous other type of surgery will be excluded from the study. Two experienced surgeons will perform all procedures; RA will perform the open and BT the laparoscopic procedures. A pylorus-resecting PD will be followed by a trans-mesocolic end-to-side hepatico-jejunostomy (HJS), a pancreatico-gastrostomy (PGS), and an ante-colic gastro-enterostomy (GES). Da Vinci robotic assistance of the reconstruction in LPD is allowed for the HJS and the PGS, while the GES will be done with endo-staplers. Statistical considerations Randomization will take place pre-operatively after informed consent has been obtained. Patients will be randomized into two groups (OPD vs. LPD) using permuted blocks of size 6. This implies that of each series of 6 consecutive patients, three patients will be randomized in each group. There are no stratification variables. Monocentric randomized trial to compare the amount of intra-operative blood loss in LPD vs. OPD. Patients expected to undergo portal vein resection/reconstruction or any simultaneous other type of surgery will be excluded from the study. Reference intra-operative blood loss levels are available from 138 patients (June 2009 - June 2012), without portal vein resection and without additional surgery. Analysis of these data reveals that these values follow a lognormal distribution; the log-transformed blood loss has a normal distribution with mean and standard deviation equal to 6.06 and 0.83, respectively. The corresponding geometric mean equals 427ml. Based on this distribution, 42.4%, 24.7% and 15.1% of the subjects are expected to have blood loss higher than 500, 750 and 1000 ml, respectively. It is assumed that the treatment (LPD) will lead to a 50% reduction of the (geometric) mean. The impact of this assumption on the distribution of the blood loss levels implies that 15.1%, 6.4% and 3.1% of the subjects in the LPD-group are expected to have a blood loss level higher than 500, 750 and 1000 ml, respectively. Based on a two-sided two-sample pooled t-test of a mean ratio with lognormal data, a total of 50 subjects is needed to detect a two-fold reduction in blood loss (with alpha set at 5%) with 90% power.