Introduction
Most of the research endeavors targeting risk factors for postoperative fistula (POPF) after distal pancreatectomy (DP) focused on the pancreatic transection method, a modifiable variable with the potential for improving fistula rates [
1,
2]. The proposed technical variants included sharp transection with handsewn closure (using mattress or fish-mouth sutures), stapled transection, transection with energy-based devices (diathermy, ultrasonic devices, with or without ligation of the main pancreatic duct), or even anastomosis of the pancreatic stump to a Roux-en-Y jejunal limb or as a pancreaticogastrostomy [
3‐
7]. Furthermore, the use of additional biologic sealants or stump reinforcement with an omental or falciform ligament patch have been investigated with mixed results [
8,
9]. Remarkably, none of these techniques have demonstrated a clear superiority over the others in randomized controlled trials [
10,
11]. Staplers and energy-based devices have been increasingly adopted in the last decade because of the more frequent use of minimally invasive approaches and the easy, fast, and reproducible mechanism of action. Recently a new type of triple-row stapler reinforced with a preloaded bioabsorbable polyglycolic acid (PGA) felt has been marketed, with preliminary data showing a decrease in the incidence and severity of POPF compared with the standard stapler and with ultrasonic devices, provided a pancreatic thickness < 17 mm [
12‐
14]. In a recent retrospective, propensity-score matched analysis of 184 patients we suggested that the use of the triple-row reinforced stapler was associated with a sharp reduction of POPF rates relative to the ultrasonic dissector group (12% versus 40%) [
15]. Under these premises, we sought to evaluate in a randomized trial whether parenchymal transection using the triple-row reinforced stapler decreases the incidence of POPF following DP compared with ultrasonic transection.
Discussion
The present randomized clinical trial of stapled versus ultrasonic transection in DP demonstrated no significant difference in POPF rates. Analysis of secondary outcomes revealed a greater incidence of abdominal collections in the ultrasonic dissection arm, although the need for percutaneous drains was comparable between groups. POPF therefore remains a clinically relevant and unsolved issue for patients undergoing elective DP, with a formation process likely independent on the surgical technique adopted for resection and closure of the pancreatic remnant. Our findings indeed resonate with previously published randomized controlled trials that did not identify an optimal transection method able to decrease POPF [
30].
To the best of our knowledge, this is the first randomized trial of a triple-row stapler reinforced with a preloaded PGA felt. Previous studies had already shown that wrapping the pancreatic stump with a PGA mesh decreased the rate of POPF [
31,
32], and triple-row stapler had been associated with less POPF compared with the double-row staplers [
33]. The Endo GIA Reinforced Reload with Tri-Staple Technology® has been available at the authors’ institution since its introduction into the market and has been employed at the surgeon’s discretion for parenchymal transection in DP. A retrospective propensity-matched analysis comparing surgical outcomes with ultrasonic dissection (HARMONIC® Focus + or Ace +) showed a significantly decreased rate of POPF in the reinforced Tri-Staple group (12% versus 40%), constituting the backbone for the present trial [
15]. As suggested by earlier studies, patients with a parenchymal thickness > 17 mm were excluded because of a very high incidence of POPF that was independent on the type of cartridge, because of stapler closure failure of parenchymal crushing [
12]. In patients who were randomized to stapled transection we gradually compressed the pancreas with the stapler for about 2–3 min, then divided the parenchyma and released the device slowly. This has been shown to help avoiding the development of POPF [
34]. Nonetheless, the choice of the stapler cartridge was left at the single surgeon’s discretion. While cartridges with closed length < 15 mm (i.e., purple) have been shown to be particularly suitable for thin pancreata (< 12 mm), in thicker glands a longer staple height has been recommended (i.e., black) although no particular cartridge has proven to outperform the others.
In the ultrasonic dissection arm, the pancreas was transected and simultaneously sealed by coaptive coagulation at the lowest vibration level. Experimental studies proved that the lateral thermal spread is limited to 0–2 mm beyond the tissue grasped within the forceps of the device [
35]. The decreased propensity for collateral thermal damage is an important putative advantage of the Harmonic scalpel, particularly when compared with other energy devices such as monopolar and bipolar diathermy, which are commonly used for pancreatic transection in DP [
30]. However, an independent association between ultrasonic transection and a slower POPF healing has been suggested by our group [
36]. Whether this depends on thermal effects has to be fully elucidated.
Analysis of factors associated with POPF suggested that BMI and the anatomic transection level play an integral role to the process. BMI is indeed a surrogate of fatty infiltration that has been shown to correlate with a complicated clinical course [
37]. Even the transection level has been widely linked to POPF, because the pancreas shape and thickness are different at the gastroduodenal artery level, at the neck, or in the body and tail [
8,
11]. Nonetheless, only intraoperative blood transfusion was independently associated with POPF on multivariable analysis. This is in accordance with a recent systematic review and meta-analysis, and might serve as a surrogate parameter for pancreatic stump ischemia [
38]. Taken together, these results emphasize the need for perioperative composite scores to predict high-risk scenarios and help establishing individualized prevention and mitigation strategies. While these tools have been derived and successfully validated in pancreatoduodenectomy [
39], previous efforts in large, multi-institutional DP series have proven elusive [
40].
The study has some limitations. First, sub-analysis of stapler cartridges was not done. The liberal use of purple or black cartridges with PGA reinforcement possibly introduced a bias, despite the compression rate and the height difference were not associated with POPF. Another limitation could be the difference in the anatomic point of parenchymal transection. Nonetheless, the point of transection was dictated by the underlying pathology, with parenchyma-sparing procedures being favored in the context of benign to low-grade neoplasms, and this parameter did not result to be a risk factor at the adjusted analysis.
In conclusion, the present randomized controlled trial of stapled transection using a PGA-reinforced triple-row stapler versus ultrasonic transection with HARMONIC® energy devices in elective DP demonstrated no significant difference in POPF rates and no substantial clinical impact on other secondary endpoints. Therefore, the optimal technique for the management of pancreatic stump in resection of the left pancreas remains unclear and warrants further investigation.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.