Risk factors for delayed gastric emptying after a 10-year retrospective study

Authors: Carlos Jiménez-Romero, Agustín de Juan Lerma, Alberto Marcacuzco Quinto, Oscar Caso Maestro, Laura Alonso Murillo, Paula Rioja Conde, Iago Justo Alonso

Abstract

Background

Delayed gastric emptying (DGE) is a frequent complication of pancreatoduodenectomy (PD) and is associated with prolonged hospital stay, readmission, increased hospital costs and decreased quality of life. However, the pathophysiology of DGE remains unclear.

Methods

This is a retrospective study of patients who underwent PD for pancreatic or periampullary tumours. All these patients were operated between January 2012 and February 2023. The patients were divided into four groups according to the development of DGE after PD: No DGE, DGE grade A, DGE grade B and DGE grade C. The groups were compared in terms of outcomes and complications. We also analysed the preoperative and perioperative risk factors for DGE development.

Results

Between January 2012 and February 2023, a total of 250 patients underwent PD. These patients were divided into four No DGE (n = 152); DGE grade A (n = 42); DGE grade B (n = 45); and DGE grade C (n = 11). The incidence of the postoperative pancreatic fistulas (POPFs) grade B/C was significantly higher in the DGE grade C group (p < .001), and the rates of post-pancreatectomy haemorrhage (p = .004) and reoperation (p < .001) were significantly higher in the DGE grade B/C groups. A significantly higher rate of grade III–IV Clavien–Dindo complications (p < .001), longer intensive care unit (p < .001) and longer hospital stays (p < .001) were observed in the DGE grade C group; and 90-day mortality (p < .001) and morbidity (p < .001) were significantly higher in the DGE grade B/C groups. Multivariate analysis demonstrated that the POPF grade B/C was a risk factor of DGE grade B/C (OR: 9.147; 95%CI: 4.125–20.281; p < .001).

Conclusions

POPF B/C is a risk factor for grade B/C DGE. Prevention of surgical complications and early treatment could contribute to the decreased incidence of DGE.

Introduction

Pancreatoduodenectomy (PD) is the standard therapy of choice for tumours in the pancreatic head and periampullary region and is associated with a mortality rate below 5% in high-volume centres [1,2]. However, the overall incidence of post-PD complications remains between 32.4% and 61.1% [3–8], including postoperative pancreatic fistula (POPF), which has an incidence of 21.6–37.7% [2,3,5,9], and delayed gastric emptying (DGE), which has an incidence of 6–57% [3,10–15]. DGE is not a lethal complication but is often associated with anxiety or depression and a significant prolongation of hospital stay, readmission, increase in hospital costs, and decreased quality of life [16,17].

DGE is defined by the International Study Group of Pancreatic Surgery (ISGPS) as the inability to return to a standard diet by the end of the first post-PD week, including prolonged nasogastric intubation, and can be divided into grades A, B and C based on the clinical course and on post-PD management [10]. The pathophysiology of DGE after PD is not clearly understood, and previous studies have attributed DGE to many causes, including surgical resection of the duodenum leading to absence of gastrointestinal hormones [18], antroduodenal congestion due to division of the left gastric vein [19], antroduodenal ischemia due to division of the right gastric artery [16], pylorospasm due to injury of the vagal nerves during lymph node dissection along the hepatoduodenal ligament [20], gastric dysrhythmia secondary to vagotomy and intraabdominal complications (pancreatic fistula, intraabdominal abscess or haemorrhage) and angulation of the reconstructed gastrointestinal tract [21]. Because of the heterogeneity of the reported studies, the nonoperative and operative risk factors for the development of DGE are unclear. This retrospective study aimed to identify the preoperative and perioperative risk factors for the development of DGE after PD.

Materials and methods

Study design and patient recruitment

This study analyses the patients who underwent PD at our institution for tumours located in the pancreatic head and periampullary regions. The period of study comprises from January 2012 to February 2023.

The patients were divided into four groups according to the absence or presence of DGE after PD: (1) patients without DGE; (2) patients with DGE grade A; (3) patients with DGE grade B; and (4) patients with DGE grade C. A flowchart of the patient selection process is shown in Figure 1. We compared the preoperative and perioperative variables and postoperative morbimortality among the four groups to analyse the risk factors for the development of clinically relevant DGE (grade B/C). Secondary endpoints included the length of intensive care unit and postoperative hospital stays and 1-, 3- and 5-year survival.

Informed consent for surgical treatment was obtained from all patients before surgery. This retrospective cohort study was approved by our institutional review board (approval number 23/080) and conducted in accordance with the Helsinki Declaration of 1975 and revision in 1983. This study was registered in a retrospective research registry (UIN research registry 9298).

The need for local clinical research ethics committee approval was waived due to the retrospective nature of the study.

Baseline data collection

The collected data included patient demographics, comorbidities, clinical presentation, laboratory parameters, diagnostic procedures, perioperative variables, histological features, postoperative morbidity/mortality and patient survival. The follow-up period ended in August 2023. None of the patients were lost to follow-up.

Surgical technique

All patients underwent PD, including cholecystectomy, division of the junction of the common hepatic duct with the common bile duct, and resection of the pancreatic head, duodenum, and approximately 15 cm of the proximal jejunum. Lymphadenectomies of the hepatoduodenal ligament, common hepatic artery and right lateral branch of the superior mesenteric artery were also performed. Pylorus preservation was performed when the tumour was distant from the duodenum, and antrectomy was performed when the tumour was close to the proximal portion of the duodenum. During pylorus-preserving pancreatoduodenectomy (PPPD), the proximal duodenum was preserved 2–3 cm distal to the pylorus, dividing the right gastric artery and the pyloric branch of the vagus nerve. Antrectomy ranging from 20 to 30% without truncal vagotomy was performed. Tumour invasion into the portomesenteric venous axis was treated with segmental resection and end-to-end anastomosis.

Each surgeon chose the anastomosis technique (with or without a ductal stent) based on their experience, preference and the presence of risk factors, such as a Wirsung duct diameter <3 mm and/or a soft consistency of the pancreas. One (classical Child’s reconstruction) or two jejunal loops (Billroth II technique) were used for pancreaticojejunostomy at the surgeon’s discretion.

Pancreaticojejunostomy was usually performed using an end-to-side duct-to-mucosa anastomosis using the two-layer technique. According to the caliber of the Wirsung duct, pancreatic texture and surgeon’s preference, pancreaticojejunostomy was performed using an external ductal stent, internal ductal stent or nonductal stent. The external stent was closed at discharge and removed 6 weeks after surgery at the outpatient clinic. Nonductal stent was performed using either duct-to-mucosa pancreaticojejunostomy or invagination pancreaticojejunostomy. An end-to-side hepaticojejunostomy was performed 15 cm distal to the pancreaticojejunostomy. Gastrojejunostomy or duodenojejunostomy (during PPPD) were routinely performed in the antecolic position approximately 60–65 cm distal from the end-to-side hepaticojejunostomy. Two Penrose or Jackson-Pratt drains were placed superior and inferior to the pancreaticojejunostomy.

Postoperative management

Prophylactic antibiotics and parenteral nutrition were administered. Intra-abdominal drains were usually removed on postoperative day (POD) 5 or 6 in the absence of a fistula, haemorrhage or infection with amylase levels <400 IU.

In the absence of POPF, the nasogastric tube was usually removed on POD 3, and sips of water were progressively administered thereafter, proceeding with a regular diet on POD 7. Prokinetics, such as erythromycin or metoclopramide, were often administered postoperatively if the patient had DGE.

Definitions

DGE was classified according to the definition approved by the ISGPS: grade A (mild), inability to tolerate solid oral intake by POD 7 and usually no vomiting; grade B (moderate), inability to tolerate solid oral intake by POD 14 with or without vomiting; and grade C (severe), inability to tolerate solid oral intake by POD 21 with or without vomiting. Grade A is not considered clinically relevant [10].

Pancreatic fistulas were classified according to the definition formulated by the ISGPS: grade A, biochemical fistulas; grade B, requirement of change in therapy or percutaneous drainage of collections; and grade C, clinical instability requiring drainage of collections or reoperations, leading to organ failure and/or mortality [22]. Post-pancreatoduodenectomy haemorrhage (PPH) was classified according to the ISGPS definition [23]. Complications were reported using the Clavien–Dindo classification [24].

Statistical analysis

Qualitative variables are expressed as absolute numbers, and relative frequencies are expressed as percentages. Associations were analysed using the Chi-squared test or Fisher’s exact test, as appropriate. Most quantitative variables did not have a normal distribution according to the Kolmogorov–Smirnov test; therefore, all quantitative variables are expressed as medians and percentiles. The relationships between quantitative variables were analysed using the Kruskal–Wallis test.

All clinically significant variables and variables with p < .05 in a univariate analysis were subsequently investigated using a binary logistic regression model to evaluate the association between variables at baseline and 90-day mortality. We also investigated their relationship with grade B and C DGE. Results are expressed as odds ratios and 95% confidence intervals. Survival analysis was performed using the Kaplan–Meier estimator and log-rank test. A p value <.05 was considered statistically significant. Statistical analyses were performed using the SPSS Statistics version 25 (SPSS Inc., Chicago, IL).

Results

A total of 250 patients underwent PD between January 2012 and February 2023. The distribution of the patients into the four groups, according to the absence or presence of DGE after PD, was as (1) patients without DGE (n = 152; 60.8%), (2) patients with DGE grade A (n = 42, 16.8%), (3) patients with DGE grade B (n = 45, 18%) and (4) patients with DGE grade C (n = 11, 4.4%). There were no significant differences in preoperative variables (demographics, comorbidities, clinical and diagnostic procedures, histological findings and laboratory findings) among the four groups, except for duodenal obstruction, which was significantly higher in the DGE grade C group than in the other groups (Table 1). We also found no significant differences in perioperative variables, including duration of surgery, blood transfusion, caliber of the Wirsung duct, pancreas consistency and surgical techniques (pylorus or non-pylorus preservation, vascular resection and intraabdominal drainage) and histological features among the groups (Table 2).

Medical complications did not differ significantly among the groups. However, wound infection was significantly more common in the DGE grade B group than in the other groups (p < .001). Concerning more serious surgical complications, the rate of grade B/C POPF was significantly higher in the DGE grade C group than in the other groups (p < .001), whereas the rates of PPH (p = .004) and reoperation (p < .001) were significantly higher in the DGE grade B and C groups than in the other groups. A significantly higher rate of grade III–IV complications, according to the Clavien–Dindo classification, was observed in the DGE grade C group (p < .001). Intensive care unit (p < .001) and hospital (p < .001) stays were significantly longer in the DGE grade C group. The 90-day morbidity (p < .001) and mortality (p < .001) were significantly higher in the DGE grade B and C groups. Adjuvant chemotherapy was administered at a significantly lower frequency in the DGE grade C group (p = .020).

The 1-, 3- and 5-year patient survival rates were similar between the groups (Table 3). Multivariate analysis demonstrated that POPF was a risk factor for the development of post-PD DGE grades B and C (OR: 9.147; 95%CI: 4.125–20.281; p < .001) (Table 4).

Discussion

The pathophysiology of DGE is not clearly understood, but there are several abnormalities involving the autonomic nervous system, smooth muscle cells, enteric neurons and interstitial cells of Cajal [24]. The surgical procedures most associated with DGE are vagotomy, Nissen fundoplication, bariatric surgery and pancreatic surgery [25,26]. Surgical resection of duodenum may cause neural and vascular alterations and reduced serum motilin stimulation [27].

Multiple nonoperative variables have been reported as risk factors for DGE after PD, including older age [14,28], male sex [13,17], BMI >35 kg/m^2^ [29], smoking history [13], coronary disease and early enteral nutrition post-PD [28], benign pathology [16] and periampullary adenocarcinoma [13]. A recent meta-analysis showed that the presence of tumours or cardiovascular diseases, hypertension, tobacco consumption, serum albumin levels and neoadjuvant chemotherapy did not increase the risk of DGE [14].

Preoperative biliary drainage has been reported as a risk factor for DGE [14,30]; however, as biliary drainage is usually performed for patients with cholangitis, the development of DGE could be attributable to the presence of cholangitis [14,31]. In our study, we did not observe significant differences in the use of preoperative biliary drainage among the groups using endoscopic retrograde cholangiopancreatography, but we did observe a tendency toward a more frequent use of percutaneous transhepatic cholangiography in the DGE grade C group owing to duodenal obstruction. Other authors [17] have not found any correlation between preoperative biliary drainage and a significantly higher incidence of DGE.

Diabetes mellitus is the most frequent cause of gastroparesis [25], but its relationship with the development of DGE post-PD is controversial. Several studies have demonstrated no change in the incidence of DGE in patients with pre-PD diabetes [14,32], whereas other studies have found a significant increase in DGE in these patients [30,31,33]. However, there is a lack of clarity regarding the distinction between pre- and post-PD diabetes between these studies [14]. We did not find significant differences among the groups in the frequency of diabetes mellitus.

The average surgical time for PD is >5.5–6 h [29,31], and higher intraoperative blood loss is significantly associated with a higher risk of DGE [14,17]. Neither variable significantly influenced the incidence of DGE in our study. The intraoperative confirmation of a Wirsung duct caliber <3 mm and soft pancreatic consistency have been reported as risk factors for POPF post-DPC [2,3,11,34]. Additionally, there is strong evidence that POPF is a risk factor for the development of DGE, with a median incidence of 23.8% [13,16,17,29,31,35–40]. We did not observe significant differences in the frequencies of the Wirsung duct caliber or pancreatic texture among the DGE groups. However, the presence of POPF significantly correlated with the incidence and severity of DGE.

Reconstructive technical modalities have been investigated to determine their relationship with the development of DGE. The pathogenesis of DGE due to pyloric preservation is attributable to pyloric, antral and duodenal ischemia [16,41], duodenal resection, ligation of the right gastric artery, and gastric dysrhythmia [18]. Several series have indicated no association between the type of reconstruction (pyloric preservation or reconstruction) and DGE [12,14,18], whereas others have reported a statistically significant higher incidence of DGE in patients who underwent pyloric preservation [42–45]. We observed no association between either modality or the development of DGE. Another study found no significant difference in the incidence of DGE between patients who underwent conventional Billroth II and those who underwent isolated Roux-en-Y reconstruction after PD [46–48]. Previous studies have indicated that the addition of a Braun jejunojejunostomy after PD without [49] or with pyloric-preserving [50] is a significant independent protective factor against DGE. Several other studies have revealed a significantly decreased rate of DGE among patients undergoing antecolic rather than retrocolic gastrojejunostomy reconstruction [31,33,38], which was attributed to less angulation, gastrojejunostomy beyond the pancreaticojejunostomy, and a more mobile jejunal loop [30]. Conversely, some authors found no difference in the incidence of DGE based on the reconstruction approach [14,51]. The antecolic position was used in all patients in this study, regardless of whether the patient underwent gastrojejunostomy or duodenojejunostomy.

Generally, postoperative complications are significantly associated with a higher incidence of DGE [29,33]. Post-PD complications such as PPH [14,17], intraabdominal collection [28,31,40], intraabdominal abscess [31,36,38,44] and reoperation [31] have been identified as severe risk factors for DGE, which occurs in up to 65% of patients with such complications [52]. In our study, POPF, PPH, reoperation and complications Clavien–Dindo grade III–IV were significantly more common in patients with DGE grade B and C. Hence, in the presence of DGE, surgeons should investigate the eventual coexistence of surgical complications as possible causes and perform early treatment that facilitates recovery from DGE [40].

In our study, the presence of surgical complications was significantly correlated with the development of DGE grade B or C, longer intensive care unit and hospital stays, and a significantly lower use of adjuvant chemotherapy due to the longer time elapsed from surgery, which is in accordance with the results of previous studies [13,30,53]. Therefore, a lower rate of surgical complications is associated with a decreased incidence of DGE. The 90-day morbidity and mortality (eight patients) were significantly higher in patients with DGE grades B and C, but actuarial patient survival was similar among the groups. All patients died of severe complications secondary to the development of grade C POPF. Our multivariate analysis demonstrated that POPF was the only risk factor for the development of DGE grade B/C, which is in line with the results of previous studies [13,16,17,31,35–40].

The retrospective nature of this study presents some limitations and makes it subject to bias. Future multicentre prospective randomized studies with large sample sizes are necessary to demonstrate the optimal technique after PD to reduce DGE.

We conclude that surgical complications, such as grade B/C POPF, grade B/C PPH and abdominal reoperation, were correlated with a significantly higher incidence of DGE grades B and C. Additionally, significantly higher rates of grade III–IV Clavien–Dindo complications, longer intensive care unit stays and longer hospital stays were observed in the DGE grade C group, whereas the 90-day mortality and morbidity were significantly higher in the DGE grade B and C groups. However, multivariate analysis demonstrated that only grade B/C POPF was a risk factor for clinically significant DGE. Prevention of surgical complications and early treatment could contribute to the decreased incidence of DGE.