Key Points
Question
Does the absence of systematic nasogastric decompression with a nasogastric tube after pancreaticoduodenectomy affect the occurrence of postoperative complications?
Findings
This randomized clinical trial of 111 patients found no evidence of a significant difference in postoperative complication occurrence (Clavien-Dindo classification grade II or higher) between systematic nasogastric decompression and no nasogastric decompression after pancreaticoduodenectomy.
Meaning
After pancreaticoduodenectomy, avoiding systematic nasogastric decompression may be considered safe.
Abstract
Importance
Although standardization of pancreaticoduodenectomy (PD) has improved postoperative outcomes, morbidity remains high. Fast-track surgery programs appear to improve morbidity, and avoiding nasogastric tube decompression (NGTD), already outdated in most major abdominal surgery, is strongly suggested after PD by fast-track surgery programs but lacks high-level evidence, especially regarding safety.
Objective
To assess in a randomized clinical trial whether the absence of systematic NGTD after PD reduces postoperative complications.
Design, Setting, and Participants
The IPOD study (Impact of the Absence of Nasogastric Decompression After Pancreaticoduodenectomy) was an open-label, prospective, single-center, randomized clinical trial conducted at a high-volume pancreatic surgery university hospital in France. In total, 155 patients who were 18 to 75 years of age and required PD for benign or malignant disease were screened for study eligibility. Key exclusion criteria were previous gastric or esophageal surgery and severe comorbidities. Patients were randomly assigned (1:1) to systematic NGTD or to no nasogastric decompression and were followed up until 90 days after surgery.
Interventions
For patients without NGTD, the NGT was removed immediately after surgery, whereas for patients with NGTD, the NGT was removed 3 to 5 days after surgery.
Main Outcomes and Measures
The primary end point was the occurrence of postoperative complications grade II or higher using the Clavien-Dindo classification. The primary end point and safety were evaluated in the intent-to-treat population.
Results
From January 2016 to August 2018, 125 screened patients were considered eligible for the study, and 111 were randomized to no NGTD (n = 52) or to NGTD (n = 59). No patient was lost to follow-up. The 2 groups had similar patient demographic and clinical characteristics at baseline. The median (interquartile range) age was 63.0 (57.0-66.5) years in the group with NGTD (38 [64.4%] were males) and 64.0 (58.0-68.0) years in the group without NGTD (31 [59.6%] were males). The postoperative complication rates grade II or higher were similar between the 2 groups (risk ratio, 0.99; 95% CI, 0.66-1.47; P > .99). Pulmonary complication rates (risk ratio, 0.59; 95% CI, 0.18-1.95; P = .44) and delayed gastric emptying rates (risk ratio, 1.07; 95% CI, 0.52-2.21; P > .99) were not significantly different between the groups. Median (interquartile) length of hospital stay for patients without NGTD was not significantly different compared with those with NGTD (10.0 [9.0-16.3] vs 12.0 [10.0-16.0] days; P = .14).
Conclusions and Relevance
The present study found no significant difference in postoperative complication occurrence of Clavien-Dindo classification grade II or higher between systematic NGTD and no NGTD after PD, suggesting that avoiding systematic nasogastric decompression is safe for this indication.
Trial Registration
ClinicalTrials.gov Identifier: NCT02594956
This open-label randomized clinical trial examines the utility and safety of nasogastric decompression after pancreaticoduodenectomy by comparing the occurrence of postoperative complications among patients with or without postoperative nasogastric tube decompression.
Introduction
Pancreaticoduodenectomy (PD) remains the only curative treatment available for pancreatic and periampullary tumor. Even with major progress in the standardization of the technique (perioperative care, anesthesiology, and surgery), PD is still associated with a high rate of postoperative morbidity, inducing prolonged hospital stay, high hospital cost, and altered quality of life.1,2,3,4 Above all, postoperative morbidity alters oncologic results owing to both delayed chemotherapy (until impossibility of realization) and promoted tumor recurrence.5
In addition to standardization and centralization of major surgical procedures, the most promising area of research is likely fast-track surgery programs (or enhanced recovery after surgery) to decrease postoperative morbidity. Among the key points of these programs6 is the lack of systematic nasogastric decompression using a nasogastric tube (NGT). In fact, an NGT itself can induce morbidity, in particular, as well as pulmonary events and increased length of hospital stay. In addition, NGT decompression (NGTD) has been abandoned in liver, gastric, and colonic surgical procedures.7,8,9
The lack of any advantage of systematic NGTD following PD has been previously suggested in a few retrospective studies10,11,12,13 although without the stronger level of evidence that is associated with the results of a randomized clinical trial (RCT). Moreover, the available data on safety from retrospective studies with historical controls may explain the reluctance by some surgeons in current practice to avoid the use of NGTD after PD.
In this context of uncertainty concerning the utility of an NGT after PD, we conducted IPOD (Impact of the Absence of Nasogastric Decompression After Pancreaticoduodenectomy), an open-label RCT comparing systematic NGTD with no NGTD after PD.
Methods
Study Design and Participants
The IPOD study was a prospective, single-center RCT with 2 groups comparing postoperative outcomes between systematic gastric decompression or no gastric decompression after PD (trial protocol in Supplement 1). The study was performed in accordance with good clinical practice guidelines and the principles of the Declaration of Helsinki.14 This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. The study protocol was approved by an independent ethics committee (CPP Ouest VI, Brest, France) October 2015. The institutional promoter was the University Hospital of Rennes, France. Written informed consent was obtained from all patients in accordance with the Declaration of Helsinki. No one received compensation or was offered any incentive for participating in this study.
All patients scheduled for PD at the University Hospital of Rennes, France, were screened and assessed for IPOD trial eligibility. Key inclusion criteria were patients between 18 and 75 years of age requiring PD for benign or malignant disease of the biliopancreatic confluence. Key exclusion criteria were previous gastric or esophageal surgery, severe comorbidities, such as end-stage kidney disease (creatinine clearance, <15 mL/min/1.73 m2; to convert to mL/s/m2, multiply by 0.0167), documented chronic respiratory disease, heart failure (New York Heart Association class III or higher), pregnancy, nursing mothers, and persons under legal protection (guardianship).
Randomization and Masking
Patients were randomly assigned (Figure) in a 1:1 ratio to the NGTD group or the no NGTD group following surgery (just before the start of parietal closure) to confirm the absence of contraindications for surgery (metastases, carcinomatosis, or arterial involvement) and to avoid influencing the procedure. The randomized group was generated for each patient the day before surgery by an independent person via a website interface, and sealed envelopes (blinded for allocation) were opened near the end of the surgical procedure, during wall closure. Because of the evident impossibility to blind the final procedure (ie, NGT placement), the study was open label without masking.
Figure. CONSORT Diagram of Patient Screening, Randomization, and Follow-up.
Procedures
Standard of Perioperative Care
Complete staging was performed in case of pancreatic cancer. Preoperative biliary drainage was performed endoscopically or radiologically if conjugated bilirubin levels reached 14.62 mg/dL (to convert to micromoles per liter, multiply by 17.104) or cholangitis was present. Undernutrition was investigated in all patients with kinetic weight loss, prealbuminemia, and albuminemia at baseline evaluation. Oral Impact Immunonutrition (Nestlé Health Science) was systematically administered to patients during the 7 days before surgery, and supplemental enteral nutrition was introduced to patients who had lost more than 10% of their previous body mass or who had severe undernutrition.
Surgery
All PD procedures were performed by senior pancreatic surgeons (K.B., L.S.) via laparotomy, full laparoscopy, or a hybrid procedure. The hybrid procedure was defined by laparoscopic resection and reconstruction through a short laparotomy. A standard lymph node dissection was performed, including the right lateral and superior aspect of the mesenteric superior artery and a paraaortic lymph node picking under the left renal vein in case of pancreatic head carcinoma. All reconstruction procedures were standardized without pylorus preservation pancreatojejunostomy and antecolic gastrojejunal anastomosis. Pancreatojejunostomy used stenting (external or internal) when the caliber of the main duct was less than 3 mm or for a soft pancreas. Vascular (vein or artery) resection was performed when necessary. Enteral feeding with a nasojejunal tube was performed according to the local practice protocol as previously published.15 Systematic drainage was performed to precisely define the postoperative pancreatic fistula according to the International Study Group of Pancreatic Fistula (ISGPF) classification.1
Anesthesia was administered according to the routine institution protocol used at the Rennes University Hospital. General anesthesia was induced with administration of intravenous (IV) propofol, sufentanil citrate, and cisatracurium and maintained with either sevoflurane or desflurane. Continuous IV ketamine hydrochloride was administered intraoperatively at antihyperalgesic doses. Patients also benefited from administration of either an epidural with ropivacaine hydrochloride for 24 to 48 hours or with continuous IV lidocaine hydrochloride intraoperatively and for 12 hours after surgery. The choice between the 2 techniques was left to the senior anesthetist in charge of the patient. Postoperative nausea and vomiting prevention was ensured with administration of IV dexamethasone sodium phosphate, 8 mg, during induction of anesthesia and then with IV droperidol or IV ondansetron hydrochloride according to the Apfel score.16 During surgery, esophageal Doppler monitoring of stroke volume was used to guide intraoperative fluid administration. For antibiotic prophylaxis, IV cefoxitin sodium, 2 g, was administered during induction of anesthesia and followed by administration of 1 g IV every 2 hours during surgery. In case of preoperative biliary drainage, patients received IV ceftriaxone, 2 g, during induction associated with metronidazole, 500 mg, adapted postoperatively according to intraoperative bacteriologic samples. Perioperative venous thromboembolism prophylaxis was adapted to patient needs according to the latest guidelines from the French Society of Anesthesiology.17
Management of NGT
In the group with NGTD, the NGT was removed on day 3 after surgery if the NGT volume was less than 500 mL or on day 5 in the absence of delayed gastric emptying. After NGT ablation, oral food intake was introduced gradually with liquid and then solid food. In the group without NGTD, the NGT was removed immediately after surgery. Liquid was introduced that night and the day after the surgery. Solid food was gradually introduced according to patient tolerance.
Postoperative Care
Postoperative analgesia was multimodal and included administration of paracetamol, nefopam, or IV lidocaine for 12 hours or epidural analgesia with ropivacaine for 24 to 48 hours and morphine patient-controlled analgesia as rescue analgesia. Mobilization was begun as soon as possible (usually the evening of the intervention). Respiratory physiotherapy was performed in all patients the first days after surgery to prevent congestion, pleural effusion, and atelectasis. Enteral feeding with a nasojejunal tube was performed according to the local practice protocol for all patients.15
Drainage was removed in the absence of hemorrhage or pancreatic fistula aspect on day 4 or 5 after surgery, depending on the amylase level ratio measurement (drain amylase level to serum amylase level <3) on days 3 and 5. In cases of pancreatic fistula or chyle leakage, drains were maintained if needed.
Outcomes
The primary end point was the occurrence of a Clavien-Dindo classification grade II or higher postoperative complication (range I to V, with higher scores indicating more severe complication). The secondary outcomes were occurrence of pulmonary complications (including atelectasis, pleural effusion, and pneumonitis); occurrence of delayed gastric emptying (based on the International Study Group on Pancreatic Surgery guidelines as A, B, or C, with increasing clinical severity18); occurrence of pancreatic fistula (classified according to the ISGPF classification as A, B, or C, with increasing clinical severity19); occurrence of biliary fistula or hemorrhage; the need for surgical reintervention; NGT reinsertion rate (time and causes); 90-day mortality rate; length of hospital stay; and the readmission rate until 90 days after surgery.
Patients were followed up until 90 days after surgery (on days 30 and 90) using standardized case report forms completed by local study coordinators and then verified by an independent clinical research associate as stated in the protocol. Preoperative data were collected by the investigator or by an independent clinical research associate as stated in the protocol (trial protocol in Supplement 1). All data required by the protocol were collected and entered in an anonymous prospective database by an independent clinical research associate as stated in the protocol.
Reinsertion of NGT and Safety
Patient safety was monitored by a board that could stop the study if adverse events occurred or more than 8 of the first 10 patients required NGT replacement. To avoid bias, surgery was performed only by 2 senior surgeons (K.B., L.S.) with substantial experience in pancreatic resection, including PD.
Statistical Analysis
Sample size calculation was based on an expected rate of postoperative complications after PD classified as Clavien-Dindo grade II or higher of 55% using systematic NGTD.2 On the basis of local preliminary results of this management performed in a sample of patients compared with a historical control, we hypothesized that this rate would decrease to 30% in the group without NGTD (absolute decrease of 25%). Under this assumption, a sample size of 50 patients per group was necessary with an α risk of 5%, a power of 80% in a 2-sided test.
Analyses were conducted with the intent-to-treat population. The primary end point was compared between the 2 groups with the χ2 test. Secondary end points were compared with t tests or Mann-Whitney tests as appropriate for continuous variables, and with χ2 tests or Fisher exact tests as appropriate for qualitative variables. For primary and secondary outcomes, risk ratios with 95% CIs were calculated for qualitative variables. Qualitative variables were expressed as number of observations (percentages), and continuous variables were expressed as mean (SD) values or as median values with interquartile ranges (IQRs). Patients who did not reach experiment end points (such as those with unplanned readmission or diet tolerance), owing to early postoperative death, were excluded from the analysis (detailed in the respective tables). We also conducted an exploratory analysis to identify risk factors for NGT reinsertion in the group without NGTD to identify risk factors that could help better identify patients who would or would not need an NGT immediately after surgery. Multivariable analysis with logistic regression was not conducted for this purpose because of the impossibility of obtaining a stable and satisfactory model.
All analyses were 2-sided, and the threshold for statistical significance was set at P < .05. Analyses were performed with R statistical software, version 3.6.3 (The R Project for Statistical Computing).
Results
Between January 2016 and August 2018, 125 patients were considered eligible for the IPOD study. Of them, 14 patients were not randomized owing to total pancreatectomy in 3 cases and perioperative contraindication to curative resection in 11 cases. In total, 111 patients were randomized to no NGTD (n = 52) or to NGTD (n = 59). All randomized patients were included in the intent-to-treat analysis according to the protocol, and no patient was lost to follow-up. Baseline patient characteristics were comparable between groups for demographic and clinical characteristics and preoperative treatment (Table 1). In brief, the median (IQR) age (63.0 [57.0-66.5] years vs 64.0 [58.0-68.0] years]) and the sex distribution (males, 38 [64.4%] vs 31 [59.6%]) were not significantly different between the NGTD groups. Of 52 patients in the no NGTD group, 25 (48.1%) had preoperative biliary drainage and 9 (17.3%) had neoadjuvant therapy compared with 29 of 59 patients (49.2%) with preoperative biliary drainage and 6 of 59 patients (10.2%) with neoadjuvant therapy in the NGTD group. Detailed perioperative information and pathologic findings are given in Table 2. The median operative time (NGTD, 325.0 [IQR, 263.5-382.5] minutes vs without NGTD, 307.5 [IQR, 250.0-380.0] minutes; P = .30), pancreatic texture (with soft pancreas: NGTD, 25 [42.4%] vs without NGTD, 29 [55.8%]; P = .22), and vein resection rate (NGTD, 13 [22.0%] vs without NGTD, 13 [25.0%]; P = .88) were not significantly different between groups. Surgical procedures were performed with laparotomy in 48 of 59 patients (81.4%) in the NGTD groups and in 45 of 52 patients (86.5%) in the group without NGTD (P = .79).
Table 1. Demographic and Clinical Characteristics of 111 Patients at Baseline.
| Characteristic | NGTD, No. (%) of patients | |
|---|---|---|
| With (n = 59) | Without (n = 52) | |
| Age, median (IQR), y | 63.0 (57.0-66.5) | 64.0 (58.0-68.0) |
| Sex | ||
| Female | 21 (35.6) | 21 (40.4) |
| Male | 38 (64.4) | 31 (59.6) |
| BMI, median (IQR) | 23.7 (21.9-26.5) | 24.1 (20.2-26.2) |
| Arterial hypertension | ||
| Yes | 13 (22.0) | 13 (25.0) |
| No | 46 (78.0) | 39 (75.0) |
| Diabetes | ||
| Yes | 9 (15.3) | 12 (23.1) |
| No | 50 (84.7) | 40 (76.9) |
| Tobacco use | ||
| Yes | 25 (42.4) | 20 (38.5) |
| No | 34 (57.6) | 32 (61.5) |
| ASA score | ||
| I | 8 (13.6) | 7 (13.5) |
| II | 37 (62.7) | 37 (71.2) |
| III | 13 (22.0) | 8 (15.4) |
| IV | 1 (1.7) | 0 |
| Preoperative biliary drainage | ||
| Yes | 29 (49.2) | 25 (48.1) |
| No | 30 (50.8) | 27 (51.9) |
| Neoadjuvant therapy | ||
| Yes | 6 (10.2) | 9 (17.3) |
| No | 53 (89.8) | 43 (82.7) |
Abbreviations: ASA, American Society of Anesthesiologists; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); IQR, interquartile range; NGTD, nasogastric tube decompression.
Table 2. Perioperative Data and Histopathologic Findings in 111 Patients.
| Variable | NGTD, No. (%) of patients | P value | |
|---|---|---|---|
| With (n = 59) | Without (n = 52) | ||
| Pain control | |||
| Epidural | 12 (2.3) | 14 (26.9) | .75 |
| Intravenous analgesia | 46 (78.0) | 37 (71.2) | |
| Surgical approach | |||
| Laparotomy | 48 (81.4) | 45 (86.5) | .79 |
| Hybrid | 7 (11.9) | 4 (7.7) | |
| Totally laparoscopic | 4 (6.8) | 3 (5.8) | |
| Operative time, median (IQR), min | 325.0 (263.5-382.5) | 307.5 (250.0-380.0) | .30 |
| Pancreatic texture | |||
| Soft | 25 (42.4) | 29 (55.8) | .22 |
| Firm | 34 (57.6) | 23 (44.2) | |
| Pancreatic duct diameter, median (IQR), mm | 3.0 (2.0-4.0) | 4.0 (3.0-5.0) | .18 |
| Wirsung duct drainage | |||
| Yes | 36 (61.0) | 26 (50.0) | .33 |
| No | 23 (39.0) | 26 (50.0) | |
| Vein resection | .88 | ||
| Yes | 13 (22.0) | 13 (25.0) | |
| No | 46 (78.0) | 39 (75.0) | |
| Hepatic artery resection | |||
| Yes | 1 (1.7) | 2 (3.8) | .60 |
| No | 58 (98.3) | 50 (96.2) | |
| Organ-associated resection | |||
| Yes | 4 (6.8) | 6 (11.5) | .51 |
| No | 55 (93.2) | 46 (88.5) | |
| Estimated operative blood loss, median (IQR), mL | 200.0 (10.0-35.0) | 187.5 (100.0-500.0) | .41 |
| Blood transfusion | |||
| Yes | 5 (8.5) | 6 (11.5) | .82 |
| No | 54 (91.5) | 46 (88.5) | |
| Time to nasogastric tube ablation, median (IQR), d | 4 (3-6) | 0 | <.001 |
| Disease type | |||
| Benign | 11 (18.6) | 10 (19.2) | >.99 |
| Malignant | 48 (81.4) | 42 (80.8) | |
| Histopathology diagnosis | |||
| Pancreatic ductal adenocarcinoma | 24 (4.7) | 25 (48.1) | .56 |
| Duodenum carcinoma | 3 (5.1) | 5 (9.6) | |
| Ampullary tumor | 4 (6.8) | 5 (9.6) | |
| Pancreatic acinar cell carcinoma | 1 (1.7) | 0 | |
| Cholangiocarcinoma | 6 (1.2) | 4 (7.7) | |
| Metastasis | 4 (6.8) | 0 | |
| Intraductal papillary mucinous neoplasm | 1 (1.7) | 1 (1.9) | |
| Neuroendocrine tumor | 5 (8.5) | 2 (3.8) | |
| Tumor size, median (IQR), mm | 2.5 (1.9-3.4) | 2.5 (2.0-3.5) | .67 |
| Histopathology resection margina | |||
| R0 | 42 (87.5) | 35 (83.3) | .85 |
| R1 | 8 (16.7) | 7 (16.7) | |
| Lymph node count, median (IQR), No. | 19 (12.5-25.0) | 18.5 (10.3-27.0) | .81 |
| Lymph node involvement | |||
| Yes | 30 (5.8) | 25 (48.1) | .91 |
| No | 29 (49.2) | 27 (51.9) | |
Abbreviations: IQR, interquartile range; NGTD, nasogastric tube decompression.
Only for patients with malignant lesions.
Primary and secondary end point results are provided in Table 3. The rate of Clavien-Dindo complications grade II or higher (primary end point) was 53.8% (28 of 52 patients) for the group without NGTD vs 54.2% (32 of 59 patients) for the NGTD group (risk ratio [RR], 0.99; 95% CI, 0.66-1.47). Reinsertion of the NGT was needed in 23.1% of patients in the group without NGTD. Delayed gastric emptying (DGE) (RR, 0.79; 95% CI, 0.47-1.32; P = .48) and DGE severity based on the International Study Group of Pancreatic Surgery consensus definition (grades B and C vs A or none: RR, 1.07; 95% CI, 0.52-2.21; P = >.99) were similar between the 2 groups. No difference between the 2 groups was noted in surgical outcomes, pulmonary complication rate (without NGTD, 2 of 52 patients [3.8%] vs with NGTD, 5 of 59 [8.5%]; RR, 0.59; 95% CI, 0.18-1.95; P = .44), surgical site infection rate (RR, 1.31; 95% CI, 0.76-2.28; P = .51), or surgical reintervention rate (RR, 0.61; 95% CI, 0.23-1.62; P = .33). Unplanned intensive care unit admission was required for 6 of 59 patients (10.2%) receiving NGTD and for 2 of 52 patients (3.8%) without NGTD (RR, 0.51; 95% CI, 0.15-1.73; P = .28).
Table 3. Primary and Secondary End Points.
| End point | NGTD, No. (%) of patients | Risk ratio (95% CI) | P value | |
|---|---|---|---|---|
| With (n = 59) | Without (n = 52) | |||
| Primary end point | ||||
| Complication of Clavien-Dindo classification grade II or higher | 32 (54.2) | 28 (53.8) | 0.99 (0.66-1.47) | >.99 |
| Secondary end point | ||||
| Pulmonary complication | 5 (8.5) | 2 (3.8) | 0.59 (0.18-1.95) | .44 |
| Nasogastric tube reinsertion | 4 (6.8) | 12 (23.1) | 1.78 (1.23-2.57) | .02 |
| Delayed gastric emptying | 17 (28.8) | 11 (21.2) | 0.79 (0.47-1.32) | .48 |
| Delayed gastric emptying of ISGPS grade B or C | 4 (6.8) | 4 (7.7) | 1.07 (0.52-2.21) | >.99 |
| Pancreatic fistula of ISGPF grade B or C | 6 (10.2) | 8 (15.4) | 1.25 (0.76-2.08) | .58 |
| Bile leakage | 5 (8.5) | 4 (7.7) | 0.94 (0.44-2.02) | >.99 |
| Surgical | ||||
| Site infection | 4 (6.8) | 6 (11.5) | 1.31 (0.76-2.28) | .51 |
| Reintervention | 7 (11.9) | 3 (5.8) | 0.61 (0.23-1.62) | .33 |
| Unplanned intensive care unit admission | 6 (10.2) | 2 (3.8) | 0.51 (0.15-1.73) | .28 |
| Length of initial hospital stay, median (IQR), d | 12.0 (10.0-16.0) | 10.0 (9.0-16.3) | NA | .14 |
| Unplanned hospital readmission (POD 90)a | 16 (28.0) | 10 (20.8) | 0.80 (0.46-1.36) | .49 |
| Mortality | ||||
| POD 30 | 1 (1.7) | 4 (7.7) | 4.54 (0.52-39.33) | .18 |
| POD 90 | 4 (6.8) | 4 (7.7) | 1.07 (0.52-2.21) | >.99 |
Abbreviations: IQR, interquartile range; ISGPF, International Study Group of Pancreatic Fistula; ISGPS, International Study Group of Pancreatic Surgery; NA, not applicable; NGTD, nasogastric tube decompression; POD, postoperative day.
Patients deceased during the initial hospitalization were excluded; consequently, 57 patients were included in the group with systematic nasogastric decompression and 48 patients were included in the group without systematic nasogastric decompression.
The median hospital length of stay was 12.0 days (IQR, 10.0-16.0 days) for patients with NGTD and 10.0 days (IQR, 9.0-16.3 days) for patients without NGTD (P = .14). The 90-day mortality rate was similar between groups (6.8% with NGTD vs 7.7% without NGTD; RR, 1.07; 95% CI, 0.52-2.21; P > .99). The 90-day readmission rate was 27.1% in the NGTD group vs 19.2% in the group without NGTD (RR, 0.77; 95% CI, 0.46-1.32; P = .45). Details about postoperative death are given in eTable 1 in Supplement 2.
Additional details about postoperative outcomes are provided in Table 4. No patient in either group experienced aspiration. Details about the patients who required NGT reinsertion in the group without NGTD are provided in eTable 2 in Supplement 2.
Table 4. Complementary Postoperative Outcomes.
| Outcome | NGTD, No. (%) of patients | P value | |
|---|---|---|---|
| With (n = 59) | Without (n = 52) | ||
| Clavien-Dindo complication grade | |||
| 0/I | 27 (45.8) | 24 (46.2) | .36 |
| II | 21 (35.6) | 18 (34.6) | |
| IIIa | 1 (1.7) | 3 (5.8) | |
| IIIb | 5 (8.5) | 2 (3.8) | |
| IVa | 3 (5.1) | 0 | |
| IVb | 0 | 1 (1.9) | |
| V | 2 (3.4) | 4 (7.7) | |
| Pancreatic fistula, ISGPF grade | |||
| Biochemical leak | 18 (30.5) | 12 (23.1) | .60 |
| Grade B | 4 (6.8) | 7 (13.5) | |
| Grade C | 2 (3.4) | 1 (1.9) | |
| Delayed gastric emptying, ISGPS grade | |||
| A | 13 (22.0) | 7 (13.5) | .52 |
| B | 3 (5.1) | 4 (7.7) | |
| C | 1 (1.7) | 0 | |
| Biliary fistula | 5 (8.5) | 4 (7.7) | >.99 |
| Postpancreatectomy hemorrhage | 9 (15.3) | 3 (5.8) | .19 |
| Time to tolerated diet, median (IQR), da | |||
| Oral liquid | 4.0 (3.0-6.0) | 1.0 (1.0-2.0) | <.001 |
| Oral solid | 5.0 (4.0-7.0) | 2.0 (1.0-3.8) | <.001 |
Abbreviations: IQR, interquartile range; ISGPF, International Study Group of Pancreatic Fistula; ISGPS, International Study Group of Pancreatic Surgery; NGTD, nasogastric tube decompression.
Because evaluation of the diet tolerance was not possible owing to early postoperative death, 1 patient was excluded from analysis in the group without systematic NGTD and 2 in the group with NGTD.
Exploratory analysis identified risk factors to help make the decision to insert an NGT or not after surgery (results reported in eTable 3 in Supplement 2). In univariate analysis, only a body mass index of 25 or greater (calculated as weight in kilograms divided by height in meters squared) was identified as a risk factor for postoperative NGT reinsertion (P < .01).
Discussion
The IPOD study was an RCT comparing no NGTD following PD with systematic NGTD. In this trial, 111 patients were randomized and the postoperative morbidity (grade II or higher using the Clavien-Dindo classification [primary end point]) was similar between the 2 groups. This trial, conducted in a high-volume center, is the first RCT, to our knowledge, to evaluate the absence of systematic NGTD safety following PD. In addition to safety, our study found that NGT reinsertion was required only in 23% of the cases without increased morbidity. The results of our study validated fast-track surgery programs advocating the absence of nasogastric suction after PD. This practice should be favored and widely extended but used with caution for patients with a body mass index greater than 25.
Despite standardization of the surgical technique, postoperative morbidity (grade II or higher using the Clavien-Dindo classification at 54%) in the present study remained important (as expected in our hypothesis of size calculation) but similar to previous reports from high-volume centers.4,20,21 This complication rate must be interpreted in regard to the high rate of vein reconstruction in both groups (25% in the no NGTD group). This is in accordance with the results of the meta-analysis published by Peng et al,22 in which postoperative morbidity was associated with significant pancreatic fistula in 12.6% of cases and delayed gastric emptying in 25.2% of patients. Contrary to our first hypothesis, pulmonary complications, including pneumonitis, pleural effusion, and atelectasis, concerned only 8.5% in the group with NGTD and 3.8% in the group without NGTD. This complication rate was lower than those previously described in high-volume centers, such as the rate reported by Tran et al in 2016.3 One can suppose that the fast-track program for postoperative recovery in both groups led to a decreased pulmonary complication rate and hindered the effect of the NGT in the present study. The benefits of implementing an enhanced recovery after surgery program or fast-track programs have already been published and summarized in a recent meta-analysis.23,24
As previously reported in another study,25 DGE was the most common cause of postoperative morbidity and, logically, for the reinsertion of the NGT. Delayed gastric emptying secondary to pancreatic fistula is the most common cause but the concept of primitive DGE existence and pathophysiology is widely debated. Gastric denervation owing to the loss of parasympathetic nerves induces a reduction in the secretion of prokinetic drugs, such as motilin,26 and consequently peristaltic contractions. As described in the present trial protocol, all procedures were performed by 2 senior pancreatic surgeons (K.B., L.S.), and systematic pylorus resection PD was performed. Contrary to most nonrandomized clinical trials, the results of the recent blinded RCT on the subject conducted by Hackert et al27 reported the nonsuperiority of pylorus-resecting PD, with a not statistically significant different rate of DGE after pylorus-resecting PD (31.2%) vs pylorus-preserving PD (25.3%). When taking into account only RCTs published on the subject, there was no significant difference between pylorus preservation or resection PD on DGE in the most recent meta-analysis.28 Our results indicating a higher risk of NGT reinsertion for patients with a body mass index of 25 or higher was also identified as a DGE risk factor in the PROPP study (Pylorus Resection or Pylorus Preservation in partial Pancreatico-Duodenectomy).27
All patients in the present trial received enteral nutrition from a second nasojejunal tube, as stated in the protocol. This practice was based on a previous retrospective study by some members of our team reporting a lower rate of DGE with enteral nutrition15 and on contradictory results in the literature. The results of a French RCT comparing enteral vs total parenteral nutrition were published in 2016 (after our study protocol was created) and suggested that systematic enteral nutrition could increase the rate of postoperative complications. To remain fully consistent with our protocol, we maintained the practice of providing enteral nutrition from a second nasojejunal tube for the duration of the present study. Recently, the meta-analysis published by Tanaka et al29 concluded that routine enteral nutrition via percutaneous enteral tube could improve postoperative outcomes. We used only enteral nutrition according to consensus guidelines and reserved the use of parenteral nutrition only for when it was necessary.30,31
As stated in the protocol, the surgical approach was left to the surgeon's discretion and was minimally invasive, with a full laparoscopic or hybrid approach in selected patients, with no significant difference in the numbers of the approaches between the 2 groups (P = .79). A recent meta-analysis of the results of prospective randomized trials comparing open and laparoscopic approaches in high-volume centers (expert in both pancreatic and laparoscopic surgery) relaunched the debate.32 The difficulties in major abdominal surgery, such as PD, are the multiple tracks for improvement and the impossibility to validate them one by one with a high level of evidence (eg, an RCT). Benchmark results are increasingly reported to evaluate surgical results and to point out the keys to care improvement to concentrate efforts. To date, benchmarking results are available for major surgery (such as upper gastrointestinal surgery33) and were recently published for pancreatic surgery.34 In addition to creating references, these data constitute tools of the utmost importance to assess the relevance of minimally invasive pancreatic surgery34 compared with the classic surgical approach. The patients in the present study who underwent a minimally invasive approach (eg, full or hybrid) induced inevitable bias, and an exploratory analysis indicated an increased rate of postoperative complications with the full laparoscopic and hybrid approaches (P = .02). However, those results were compensated for by a good balance of approaches between groups owing to randomization.
Limitations
This trial does have some limitations. The limited number of patients associated with the complexity of the conditions among those patients (multiple known and unknown characteristics influencing outcome) must be considered. However, none of the patients was lost to follow-up, and only 2 experienced senior surgeons conducted the procedures using a standard method in a short and recent period with a strict postoperative management protocol that limited bias and reinforced our results, in particular vis-à-vis safety. However, our negative results with an a priori superiority hypothesis must be interpreted with pragmatism and caution. The conclusion that both attitudes are equivalent is not permitted in this setting even if we were not able to highlight a significant difference in the primary outcome. We noted that with a 5% difference in the postoperative complication rates of Clavien-Dindo grade II or higher as the cutoff, 1237 patients in each group would have needed to be included to show noninferiority between the 2 practices. Finally, we did not conduct a quality-of-life analysis to evaluate patient reactions and the influence of the procedure on their return to satisfactory functional recovery.
Conclusions
Our study found no significant difference in postoperative complication (Clavien-Dindo classification grade II or higher) occurrence between systematic NGTD and no NGTD after pancreaticoduodenectomy, suggesting that avoiding systematic nasogastric decompression may be safe for this indication.
Trial Protocol
eTable 1. Details About Postoperative Deaths
eTable 2. Detailed Characteristics of Patients Requiring Secondary Nasogastric Decompression in the NGT Group
eTable 3. Risk Factor Analysis (Univariate Analysis) for NGT Reinsertion in the NGT Group (Exploratory Analysis)
Data Sharing Statement
References
- 1.Bassi C, Dervenis C, Butturini G, et al. ; International Study Group on Pancreatic Fistula Definition . Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery. 2005;138(1):8-13. doi: 10.1016/j.surg.2005.05.001 [DOI] [PubMed] [Google Scholar]
- 2.Welsch T, Borm M, Degrate L, Hinz U, Büchler MW, Wente MN. Evaluation of the International Study Group of Pancreatic Surgery definition of delayed gastric emptying after pancreatoduodenectomy in a high-volume centre. Br J Surg. 2010;97(7):1043-1050. doi: 10.1002/bjs.7071 [DOI] [PubMed] [Google Scholar]
- 3.Tran TB, Dua MM, Worhunsky DJ, Poultsides GA, Norton JA, Visser BC. The first decade of laparoscopic pancreaticoduodenectomy in the United States: costs and outcomes using the Nationwide Inpatient Sample. Surg Endosc. 2016;30(5):1778-1783. doi: 10.1007/s00464-015-4444-y [DOI] [PubMed] [Google Scholar]
- 4.Schneider EB, Hyder O, Wolfgang CL, et al. Provider versus patient factors impacting hospital length of stay after pancreaticoduodenectomy. Surgery. 2013;154(2):152-161. doi: 10.1016/j.surg.2013.03.013 [DOI] [PubMed] [Google Scholar]
- 5.Nagai S, Fujii T, Kodera Y, et al. Recurrence pattern and prognosis of pancreatic cancer after pancreatic fistula. Ann Surg Oncol. 2011;18(8):2329-2337. doi: 10.1245/s10434-011-1604-8 [DOI] [PubMed] [Google Scholar]
- 6.Lassen K, Coolsen MME, Slim K, et al. ; Enhanced Recovery After Surgery (ERAS) Society, for Perioperative Care; European Society for Clinical Nutrition and Metabolism (ESPEN); International Association for Surgical Metabolism and Nutrition (IASMEN) . Guidelines for perioperative care for pancreaticoduodenectomy: Enhanced Recovery After Surgery (ERAS®) Society recommendations. World J Surg. 2013;37(2):240-258. doi: 10.1007/s00268-012-1771-1 [DOI] [PubMed] [Google Scholar]
- 7.Nelson R, Edwards S, Tse B. Prophylactic nasogastric decompression after abdominal surgery. Cochrane Database Syst Rev. 2007;(3):CD004929. doi: 10.1002/14651858.CD004929.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Pessaux P, Regimbeau J-M, Dondéro F, Plasse M, Mantz J, Belghiti J. Randomized clinical trial evaluating the need for routine nasogastric decompression after elective hepatic resection. Br J Surg. 2007;94(3):297-303. doi: 10.1002/bjs.5728 [DOI] [PubMed] [Google Scholar]
- 9.Carrère N, Seulin P, Julio CH, Bloom E, Gouzi J-L, Pradère B. Is nasogastric or nasojejunal decompression necessary after gastrectomy? a prospective randomized trial. World J Surg. 2007;31(1):122-127. doi: 10.1007/s00268-006-0430-9 [DOI] [PubMed] [Google Scholar]
- 10.Kunstman JW, Klemen ND, Fonseca AL, Araya DL, Salem RR. Nasogastric drainage may be unnecessary after pancreaticoduodenectomy: a comparison of routine vs selective decompression. J Am Coll Surg. 2013;217(3):481-488. doi: 10.1016/j.jamcollsurg.2013.04.031 [DOI] [PubMed] [Google Scholar]
- 11.Fisher WE, Hodges SE, Cruz G, et al. Routine nasogastric suction may be unnecessary after a pancreatic resection. HPB (Oxford). 2011;13(11):792-796. doi: 10.1111/j.1477-2574.2011.00359.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kleive D, Sahakyan MA, Labori KJ, Lassen K. Nasogastric tube on demand is rarely necessary after pancreatoduodenectomy within an enhanced recovery pathway. World J Surg. 2019;43(10):2616-2622. doi: 10.1007/s00268-019-05045-4 [DOI] [PubMed] [Google Scholar]
- 13.Gaignard E, Bergeat D, Courtin-Tanguy L, et al. Is systematic nasogastric decompression after pancreaticoduodenectomy really necessary? Langenbecks Arch Surg. 2018;403(5):573-580. doi: 10.1007/s00423-018-1688-8 [DOI] [PubMed] [Google Scholar]
- 14.World Medical Association World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191-2194. doi: 10.1001/jama.2013.281053 [DOI] [PubMed] [Google Scholar]
- 15.Rayar M, Sulpice L, Meunier B, Boudjema K. Enteral nutrition reduces delayed gastric emptying after standard pancreaticoduodenectomy with child reconstruction. J Gastrointest Surg. 2012;16(5):1004-1011. doi: 10.1007/s11605-012-1821-x [DOI] [PubMed] [Google Scholar]
- 16.Apfel CC, Läärä E, Koivuranta M, Greim CA, Roewer N. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between two centers. Anesthesiology. 1999;91(3):693-700. doi: 10.1097/00000542-199909000-00022 [DOI] [PubMed] [Google Scholar]
- 17.Samama C-M, Gafsou B, Jeandel T, et al. ; French Society of Anaesthesia and Intensive Care . Guidelines on perioperative venous thromboembolism prophylaxis: update 2011: short text [in French]. Ann Fr Anesth Reanim. 2011;30(12):947-951. Published correction appears in Ann Fr Anesth Reanim. 2012;31(1):93. doi: 10.1016/j.annfar.2011.10.008 [DOI] [PubMed] [Google Scholar]
- 18.Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142(5):761-768. doi: 10.1016/j.surg.2007.05.005 [DOI] [PubMed] [Google Scholar]
- 19.Bassi C, Marchegiani G, Dervenis C, et al. ; International Study Group on Pancreatic Surgery (ISGPS) . The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;161(3):584-591. doi: 10.1016/j.surg.2016.11.014 [DOI] [PubMed] [Google Scholar]
- 20.Witzigmann H, Diener MK, Kienkötter S, et al. No need for routine drainage after pancreatic head resection: the dual-center, randomized, controlled PANDRA trial (ISRCTN04937707). Ann Surg. 2016;264(3):528-537. doi: 10.1097/SLA.0000000000001859 [DOI] [PubMed] [Google Scholar]
- 21.van Rijssen LB, Koerkamp BG, Zwart MJ, et al. ; Dutch Pancreatic Cancer Group . Nationwide prospective audit of pancreatic surgery: design, accuracy, and outcomes of the Dutch Pancreatic Cancer Audit. HPB (Oxford). 2017;19(10):919-926. doi: 10.1016/j.hpb.2017.06.010 [DOI] [PubMed] [Google Scholar]
- 22.Peng C, Zhou D, Meng L, et al. The value of combined vein resection in pancreaticoduodenectomy for pancreatic head carcinoma: a meta-analysis. BMC Surg. 2019;19(1):84. doi: 10.1186/s12893-019-0540-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Williamsson C, Karlsson N, Sturesson C, Lindell G, Andersson R, Tingstedt B. Impact of a fast-track surgery programme for pancreaticoduodenectomy. Br J Surg. 2015;102(9):1133-1141. doi: 10.1002/bjs.9856 [DOI] [PubMed] [Google Scholar]
- 24.Ji H-B, Zhu W-T, Wei Q, Wang X-X, Wang H-B, Chen Q-P. Impact of enhanced recovery after surgery programs on pancreatic surgery: a meta-analysis. World J Gastroenterol. 2018;24(15):1666-1678. doi: 10.3748/wjg.v24.i15.1666 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Eshuis WJ, van Eijck CHJ, Gerhards MF, et al. Antecolic versus retrocolic route of the gastroenteric anastomosis after pancreatoduodenectomy: a randomized controlled trial. Ann Surg. 2014;259(1):45-51. doi: 10.1097/SLA.0b013e3182a6f529 [DOI] [PubMed] [Google Scholar]
- 26.Naritomi G, Tanaka M, Matsunaga H, et al. Pancreatic head resection with and without preservation of the duodenum: different postoperative gastric motility. Surgery. 1996;120(5):831-837. doi: 10.1016/S0039-6060(96)80091-2 [DOI] [PubMed] [Google Scholar]
- 27.Hackert T, Probst P, Knebel P, et al. Pylorus resection does not reduce delayed gastric emptying after partial pancreatoduodenectomy: a blinded randomized controlled trial (PROPP study, DRKS00004191). Ann Surg. 2018;267(6):1021-1027. doi: 10.1097/SLA.0000000000002480 [DOI] [PubMed] [Google Scholar]
- 28.Klaiber U, Probst P, Strobel O, et al. Meta-analysis of delayed gastric emptying after pylorus-preserving versus pylorus-resecting pancreatoduodenectomy. Br J Surg. 2018;105(4):339-349. doi: 10.1002/bjs.10771 [DOI] [PubMed] [Google Scholar]
- 29.Tanaka M, Heckler M, Mihaljevic AL, et al. Meta-analysis of effect of routine enteral nutrition on postoperative outcomes after pancreatoduodenectomy. Br J Surg. 2019;106(9):1138-1146. doi: 10.1002/bjs.11217 [DOI] [PubMed] [Google Scholar]
- 30.Weimann A, Braga M, Carli F, et al. ESPEN guideline: clinical nutrition in surgery. Clin Nutr. 2017;36(3):623-650. doi: 10.1016/j.clnu.2017.02.013 [DOI] [PubMed] [Google Scholar]
- 31.Arends J, Bachmann P, Baracos V, et al. ESPEN guidelines on nutrition in cancer patients. Clin Nutr. 2017;36(1):11-48. doi: 10.1016/j.clnu.2016.07.015 [DOI] [PubMed] [Google Scholar]
- 32.Nickel F, Haney CM, Kowalewski KF, et al. Laparoscopic versus open pancreaticoduodenectomy: a systematic review and meta-analysis of randomized controlled trials. Ann Surg. 2020;271(1):54-66. doi: 10.1097/SLA.0000000000003309 [DOI] [PubMed] [Google Scholar]
- 33.Low DE, Kuppusamy MK, Alderson D, et al. Benchmarking complications associated with esophagectomy. Ann Surg. 2019;269(2):291-298. doi: 10.1097/SLA.0000000000002611 [DOI] [PubMed] [Google Scholar]
- 34.Sánchez-Velázquez P, Muller X, Malleo G, et al. Benchmarks in pancreatic surgery: a novel tool for unbiased outcome comparisons. Ann Surg. 2019;270(2):211-218. doi: 10.1097/SLA.0000000000003223 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol
eTable 1. Details About Postoperative Deaths
eTable 2. Detailed Characteristics of Patients Requiring Secondary Nasogastric Decompression in the NGT Group
eTable 3. Risk Factor Analysis (Univariate Analysis) for NGT Reinsertion in the NGT Group (Exploratory Analysis)
Data Sharing Statement