Table 3.

Effects of surgical interventions on DGE following pancreatectomy

Author	Type of study	Total no. of patients	Type of intervention	Results
Yeo et al. (1999) [104]	Prospective randomized trial	114 • 56 • 58	Standard PD with enbloc removal of peri-pancreatic lymph nodes vs. radical PD with standard resection along with a distal gastrectomy and retroperitoneal lymphadenectomy	• Higher incidence of DGE in patients with radical resection (16 vs. 4 %, p = 0.03)
Yeo et al. (2002) [105]	Randomized controlled trial	299 • 146 • 148	Standard PD vs. radical (standard plus distal gastrectomy and extended retroperitoneal lymph node resection)	• Higher rates of DGE (16 vs. 6 %, p = 0.006) and POPF (13 vs. 6 %, p = 0.05) in the radical group
Tamandl et al. (2014) [110]	Prospective randomized study	64 • 36 • 28	Antecolic vs. retrocolic reconstruction of the duodenojejunostomy after pylorus-preserving PD	• Incidence of DGE on POD 10 17.6 % in the antecolic group vs. 23.1 % in the retrocolic group. (p = 0.628) • No significant difference between the two approaches
Kurahara et al. (2011) [86]	Prospective randomized controlled study	46 • 22 • 24	Retrocolic vs. antecolic reconstruction after pancreatoduodenectomy	• Significantly higher incidence of DGE in the retrocolic group compared to antecolic reconstruction (50 % vs. 20.8 %, p = 0.0364). • Higher incidence of grade B/C DGE in the retrocolic group (27.3 vs. 4.2 %, p = 0.0234)
Kawai et al. (2011) [84]	Prospective randomized controlled trial	130 • 66 • 64	Pylorus ring resection vs. pylorus preservation during pancreatoduodenectomy	• Significantly decreased incidence of DGE in intervention group (4.5 vs. 17.2 %, p = 0.0244) • Significant delay in C-acetate breath test at 1, 3, and 6 months postoperatively in pylorus-preserving PD
Matsumoto et al. (2014) [85]	Prospective randomized comparison	100 • 50 • 50	Pylorus-preserving pancreatoduodenectomy (PPPD) vs. subtotal stomach-preserving pancreatoduodenectomy (SSPPD)	• No significant difference in incidence of DGE between PPPD and SSPPD (20 vs. 12 %, p = 0.414)
Mack et al. (2004) [103]	Prospective, randomized controlled trial	36 • 20 • 16	Gastric decompression and enteral feeding by a double-lumen gastrojejunostomy tube	• 25 % controls had prolonged gastroparesis as compared to none in the intervention group (p = 0.03) • Significantly longer hospital stay in controls (p = 0.01)
Tani et al. (2006) [80]	Prospective, randomized controlled trial	40 • 20 • 20	Antecolic vs. retrocolic duodeno-jejunostomy during pylorus-preserving pancreaticoduodenectomy	• Significantly high incidence of DGE in retrocolic approach as compared to antecolic (5 vs. 50 %, p = 0.0014) • Significantly shorter duration of postoperative NG drainage, early ability to take solid foods, and significantly shorter hospital stay in patients with antecolic route
Tien et al. (2009) [111]	Prospective randomized trial	247 • 123 • 124	Modified Roux-en-Y gastrojejunostomy reconstruction with placement of a jejunostomy feeding tube vs. conventional gastric bypass (control group)	• No significant difference in incidence of DGE between the two groups • Grades of DGE were significantly lower in the modified group as compared to the control group (p = 0.01)
Shimoda et al. (2013) [99]	Prospective randomized trial	101 • 52 • 49	Billroth II vs. Roux-en-Y reconstruction for the gastrojejunostomy during subtotal stomach-preserving pancreaticoduodenectomy	• Significantly lower incidence of DGE in the Billroth group vs. the R-Y group (5.7 vs. 20.4 %, p = 0.028) • Significantly shorter duration of hospital stay in the Billroth group
Imamura et al. (2014) [112]	Prospective randomized clinical trial	116 • 58 • 58	Antecolic vs. vertical retrocolic duodo-jejunostomy during pylorus-preserving pancreaticoduodenectomy	• No significant difference between the two groups in incidence of DGE. • Better weight recovery in the vertical retrocolic group at 12 months post operatively
Gangavatiker et al. (2011) [66]	Randomized controlled trial	72 • 35 • 37 (68 included in final analysis)	Antecolic vs. retrocolic gastro/duodeno-jejunostomy during pancreaticoduodenectomy (Whipple/pylorus-preserving PD)	• No significant difference in the incidence of DGE between the two groups • Age significantly associated with occurrence of DGE
Tran et al. (2004) [95]	Prospective, randomized multicenter analysis	170 • 83 • 87	Pylorus preserving pancreaticoduodenectomy vs. standard Whipple procedure	• No significant difference in operation time, volume of blood loss, length of hospital stay, mortality, morbidity, and incidence of DGE between the two groups
Eshuis et al. (2014) [81]	Randomized control trial	246 • 125 • 121	Antecolic vs. retrocolic route of gastroenteric anastomosis after PD	• No significant difference in incidence of DGE in retrocolic group and antecolic group (45 (36 %) vs. 41 (34 %); p = 0.89)
Srinarmwong et al. (2008) [113]	Randomized control trial	27	Standard Whipple’s vs. PPPD	• DGE was more common in patients undergoing PPPD
Wellner et al. (2012) [27]	Randomized controlled trial	116 • 59 • 57	Pancreaticogastrostomy vs. pancreaticojejunostomy after partial PD	• DGE (27 vs. 17 %, p = 0.246) and intraluminal bleeding (7 vs. 2 %, p = 0.364) more frequent with PG but no statistically significant difference