Postoperative nutritional support after pancreaticoduodenectomy in adults.

BACKGROUND

Resection of the head of the pancreas is most commonly done by a pancreaticoduodenectomy, known as a Whipple procedure. The most common indication for pancreaticoduodenectomy is malignancy, but can include benign tumours and chronic pancreatitis. Complete surgical resection, with negative margins, provides the best prospect of long-term survival. Pancreaticoduodenectomy involves specific and unique alterations to the digestive system and maintaining nutritional status (optimising outcomes and achieving resumption of a normal diet) in patients with cancer after major surgery is a challenge. Malnutrition is a risk factor following pancreaticoduodenectomy, due to the magnitude of the operation and the frequency of complications. Postoperatively, patients are fed either orally, enterally or parenterally. Oral intake may start with fluids and then progress to solid food, or may be ad libitum. Enteral feeding may be via a nasojejunal tube or feeding tube jejunostomy. Parenteral nutrition can be delivered via a central or peripheral intravenous line, and may provide full nutrition (TPN) or partial nutrition (supplemental PN).

OBJECTIVES

To assess the effects of postoperative nutritional support strategies on complications and recovery in adults after pancreaticoduodenectomy.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, LILACS and CINAHL (from inception to October 2022), ongoing trials registers and other internet databases. We searched previous systematic reviews, relevant publications on the same topic and the references of included studies.

SELECTION CRITERIA

Randomised controlled trials of postoperative nutritional interventions in an inpatient setting for patients undergoing pancreaticoduodenectomy. We specifically looked for studies comparing route or timing rather than nutritional content.

DATA COLLECTION AND ANALYSIS

Two review authors independently assessed studies for inclusion, judged the risk of bias and extracted data. Studies requiring translation were assessed for inclusion, risk of bias and data extraction by an external translator and another author. We used GRADE to evaluate the certainty of the evidence.

MAIN RESULTS

We included 17 studies (1897 participants). Of these, eight studies could be included in a meta-analysis. The route, timing and target of nutritional support varied widely between studies. Enteral feeding (jejunostomy, nasojejunal or gastrojejunostomy) was used in at least 13 studies (one study did not specify the method of enteral route), parenteral nutrition (PN) was used in at least 10 studies (two studies had a control of 'surgeon's preference' and no further details were given) and oral intake was used in seven studies. Overall, the evidence presented in this review is of low to very low certainty. Four studies compared jejunostomy feeding with total parenteral nutrition. When we pooled these four studies, the evidence demonstrated that jejunostomy likely results in a reduced length of hospital stay (mean difference (MD) -1.61 days, 95% confidence interval (CI) -2.31 to -0.92; 3 studies, 316 participants; moderate-certainty evidence). The evidence suggested that there may be no difference in postoperative pancreatic fistula (risk ratio (RR) 0.77, 95% CI 0.41 to 1.47; 4 studies, 346 participants; low-certainty evidence) and that there may be no difference in delayed gastric emptying (RR 0.38, 95% CI 0.04 to 3.50; 2 studies, 270 participants; very low-certainty evidence) or post pancreatectomy haemorrhage (RR 0.36, 95% CI 0.06 to 2.29; 2 studies, 270 participants; very low-certainty evidence), but the evidence is uncertain. There were no data for major and minor complications defined by the Clavien-Dindo classification. Two studies compared nasojejunal feeding with total parenteral nutrition. When the two studies were pooled, the evidence suggested that there may be little to no difference between nasojejunal feeding and TPN in the length of hospital stay (MD 1.07 days, 95% CI -2.64 to 4.79; 2 studies, 242 participants; low-certainty evidence), delayed gastric emptying (RR 1.26, 95% CI 0.83 to 1.91; 2 studies, 242 participants; low-certainty evidence) or post pancreatectomy haemorrhage (RR 1.00, 95% CI 0.62 to 1.62; 2 studies, 242 participants; low-certainty evidence). TPN may slightly improve rates of clinically relevant postoperative pancreatic fistula (RR 2.13, 95% CI 1.21 to 3.74; 2 studies, 242 participants; low-certainty evidence). One study reported on major complications (RR 1.27, 95% CI 0.83 to 1.94; very low-certainty evidence) and minor complications (RR 1.01, 95% CI 0.68 to 1.50; 204 participants; very low-certainty evidence) defined by the Clavien-Dindo classification and there may be little to no difference in effect, but the evidence is uncertain. Two studies compared jejunostomy feeding with oral intake. Of note, one of the studies used a modified surgical technique as part of the intervention. We pooled these studies and found that there may be little to no difference in the length of hospital stay (MD -1.99 days, 95% CI -4.90 to 0.91; 2 studies, 301 participants; very low-certainty evidence) or delayed gastric emptying (RR 0.98, 95% CI 0.33 to 2.88; 2 studies, 307 participants; very low-certainty evidence). One study reported on major complications (RR 1.01, 95% CI 0.44 to 2.34; 247 participants; very low-certainty evidence) and minor complications (RR 0.83, 95% CI 0.59 to 1.15; 247 participants; very low-certainty evidence) defined by the Clavien-Dindo classification, postoperative pancreatic fistula (RR 0.86, 95% CI 0.30 to 2.50; 247 participants; very low-certainty evidence) and post pancreatectomy haemorrhage (RR 2.02, 95% CI 0.52 to 7.88; 247 participants; very low-certainty evidence) and there may be little to no difference in effect on these outcomes, but the evidence is uncertain. No difference in mortality was detected in any of the analyses (Clavien-Dindo Grade V) (very low-certainty evidence).

AUTHORS' CONCLUSIONS: When compared with parenteral nutrition, enteral nutrition by jejunostomy likely results in a decreased length of hospital stay and may lead to no difference in the incidence of postoperative complications. When compared with parenteral nutrition, enteral feeding by nasojejunal tube may result in no difference in the incidence of postoperative complications or length of hospital stay. When compared with oral nutrition, enteral nutrition by jejunostomy feeding may result in no difference in the incidence of postoperative complications or length of hospital stay, but the evidence is very uncertain. Further high-quality research is required and there are several ongoing studies. Given the number of different nutritional interventions available in the postoperative setting, a network meta-analysis would be more appropriate in future.