Impact of foregoing a nasogastric tube in robotic pancreaticoduodenectomy: a propensity score-matched study

doi:10.21203/rs.3.rs-4369948/v1

Download PDF

Research Article

Impact of foregoing a nasogastric tube in robotic pancreaticoduodenectomy: a propensity score-matched study

https://doi.org/10.21203/rs.3.rs-4369948/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Nasogastric tube (NGT) placement has been standard practice in pancreatoduodenectomy, but also questioned. This study evaluated the feasibility and safety of foregoing a NGT in robotic pancreatoduodenectomy (RPD). A total of 906 patients undergoing RPD or open pancreatoduodenectomy (OPD) were included in this study for propensity score matching. Study group of RPD without a NG tube (RPD(-)NGT) was compared with those of RPD with a NG tube (RPD(+)NGT) and OPD with a NG tube (OPD(+)NGT). Each arm comprised 100 patients. The RPD groups without and with NGT had less intraoperative blood loss compared to the OPD group, with a median of 155 vs. 200 vs. 500 cc (p < 0.001) and a higher lymph node yield, with a median number of 18 vs. 15 vs. 12 (p < 0.001). The surgical outcomes were comparable among the RPD(-)NGT, RPD(+)NGT, and OPD(+)NGT groups, except for delayed gastric emptying (DGE) and length of stay (LOS). The incidence of DGE was significantly lower in the RPD without and with NGT groups than in the OPD(+)NGT group (3% vs. 4% vs. 18%, p < 0.001). The LOS was also shorter in the RPD(-)NGT group, with a median of 17 days, compared to the other two groups with NGT, 22 days and 26 days (p < 0.001). Only operation type, OPD, remained an independent risk factor for predicting the occurrence of DGE. RPD without NGT was associated with lower DGE and shorter LOS. It is feasible to forego a NGT in RPD cases without negative impact on safety.

nasogastric tube

delayed gastric emptying

pancreaticoduodenectomy

robotic

Nasogastric tube (NGT) decompression has traditionally been a routine practice following most major gastrointestinal surgeries because it has been thought that the use of a NGT after major intra-abdominal procedures could reduce postoperative complications, such as vomiting, wound dehiscence, and anastomotic leaks [1, 2]. Proponents of NGT decompression also believe that it may help prevent distension, potentially lowering the risk of wound dehiscence, and facilitating a quicker return of bowel function [1, 3]. Pancreaticoduodenectomy, one of the most challenging abdominal surgeries, has historically carried a high morbidity rate ranging from 40–50%. Advances in surgical techniques and the concentration of cases in high-volume centers have resulted in a decreased mortality rate of less than 5% [4, 5]. Postoperative pancreatic fistula (POPF) is the most feared complication following pancreaticoduodenectomy, while delayed gastric emptying (DGE) is a common postoperative issue after open pancreaticoduodenectomy (OPD), with an incidence ranging from 4.5–56.1% of patients [5–8]. This is often associated with a protracted hospitalization [6]. Owing to the heightened risk of DGE, many pancreatic surgeons would prefer to use NGT for postoperative nasogastric decompression following pancreaticoduodenectomy. Placement of a NGT after pancreatoduodenectomy has been the standard procedure in most medical centers for several decades [9].

Some researchers question the routine placement of NGT following pancreatoduodenectomy [6]. In fact, certain studies have presented contrasting findings, suggesting that foregoing the use of a NGT can expedite the return of bowel function, reduce pulmonary complications, and not lead to an increase in anastomotic leaks [2]. Despite these findings, many surgeons remain hesitant to abandon the routine practice of postoperative NGT placement owing to the high incidence of DGE and the intricate nature of enteric anastomoses in pancreatoduodenectomy. Several studies have reported that routine use of NGT after pancreatoduodenectomy is unnecessary [2, 3, 9–13]. The advent of the da Vinci Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA, USA) has brought about a significant transformation in the field of minimally invasive surgery (MIS), revolutionizing various surgical procedures, including complex pancreatic surgery. One notable benefit of this technology is the achievement of smaller incisions, which result in reduced wound pain and improved cosmetic outcomes. Furthermore, our previous study demonstrated favorable surgical outcomes with a low DGE in robotic pancreatoduodenectomy (RPD) [14].

This study aimed to evaluate the feasibility and safety of not using a NGT after RPD. We assessed the DGE and length of stay (LOS) in three different groups: one study group of RPD without NG tube (RPD(-)NGT), and two control groups of RPD with NG tube (RPD(+)NGT) and OPD with NG tube (OPD(+)NGT). All procedures were performed by the same surgical team using the same technique. Additionally, this study aimed to compare the surgical outcomes among these three groups.

Based on our previous experience of a low incidence of DGE in RPD [5], RPD without NGT placement has become our standard of care since March 2022. The NGT is routinely removed in the operating room immediately after intratracheal extubation. Before March 2023, NGT placement was part of our routine practice after pancreaticoduodenectomy, including RPD and OPD. For this study, we identified 906 patients with periampullary lesions who underwent RPD or OPD between July 2012 and May 2023. The patients were selected from a prospectively collected computer database and divided into three groups: one study group of RPD without NG tube (RPD(-)NGT), and two control groups of RPD with NG tube (RPD(+)NGT) and OPD with NG tube (OPD(+)NGT). This study was approved by the Institutional Review Board (IRB) of Taipei Veterans General Hospital (IRB-TPEVGH No. 2023-07-007CC) and followed the ethical guidelines for the human studies proposed by our IRB. The requirement for informed consent was waived for this retrospective cohort study, with data anonymity. The exclusion criteria for RPD included long vascular encasements (> 2 cm) and severe intraabdominal adhesions. Otherwise, patient selection for RPD was based on individual preferences after detailed counseling about the innovative nature of RPD, its advantages, disadvantages, and the availability of a robotic machine. Demographic and clinical variables, including sex, age, body mass index (BMI), jaundice, diabetes mellitus (DM), American Society of Anesthesiologists (ASA) physical status classification, clinical presentation, and diagnosis, were assessed. Pathological variables, such as malignancy, tumor size, lymph node status, perineural invasion, lymphovascular invasion, and tumor stage, were also included in the evaluation. Intraoperative variables for evaluation included operation type (RPD vs. OPD), pylorus-preserving pancreaticoduodenectomy (PPPD), operation time, mesopancreas dissection levels, blood loss, vascular resection, and tumor radicality. Postoperative surgical outcomes included surgical mortality, various complications (including DGE), and LOS. The discharge criteria after pancreaticoduodenectomy at our institute included stable vital signs without fever after removal of the drain, unless patients could take care of themselves. Because our national health insurance system paid the hospital charge, our patients after major operation usually did not leave hospital early.

Study endpoints

This study aimed to investigate the feasibility and safety of not placing a NGT in patients undergoing RPD. The primary endpoint was to assess the feasibility by comparing the rates of DGE and LOS among the study group of RPD(-)NGT and the two control groups of RPD(+) NG and OPD(+)NGT groups. The secondary endpoint was to evaluate safety by comparing the surgical outcomes between the study group of RPD(-)NGT group and the two control groups with NGT.

Propensity score matching (PSM) strategy

PSM was conducted to mitigate selection bias when estimating the causal treatment effects and to achieve a balance in potential confounders between the study group of RPD(-)NGT and control groups of NGT (RPD(+)NGT and OPD(+)NGT). An individual propensity score was derived through logistic regression modeling, utilizing the following commonly used six covariates to predict the occurrence of DGE: age, jaundice, DM, malignancy, lymph node metastasis, and PPPD [5]. Two steps of PSM were subsequently performed; first step was the PSM between the study group without NGT, RPD(-)NGT and the control group with RPD(+)NGT in a 1:1 paring ratio, with the matches starting from cases with the largest propensity score; the second step was the other PSM between the study group without NGT, RPD(-)NGT and the other control group with OPD(+)NGT in a 1:1 paring ratio. To increase the need for sample size, a specific caliper width of 0.02 standard deviations of the logit of the estimated propensity score was applied.

Surgical technique

All surgical procedures were performed by the same surgical team using the same technique, either via a robotic or open approach. The technique of pancreaticoduodenectomy with modified Blumgart pancreaticojejunostomy reconstruction has been described in detail previously [4, 15]. RPD was performed using the Si or Xi da Vinci Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA, USA). In the RPD group, the Harmonic® scalpel, an energy device, was used for dividing the small vessels, and Hem-o-lok® systems (Teleflex Inc., Chelmsford, MA, USA) were selectively used for the large vessels. In the OPD group, most vascular pedicles were cauterized or selectively ligated. The right gastric artery was routinely divided. If feasible, PPPD was attempted; otherwise, limited antrectomy was performed for patients with an ischemic pylorus after dividing the right gastric artery.

In the RPD group, the final anastomosis was a handsewn gastrojejunostomy performed using an extracorporeal approach with careful downward positioning of the stomach. The gastrojejunostomy tube was positioned in an antecolic, antiperistaltic, and inframesocolic position near the umbilicus, resulting in a relatively vertical position of the stomach after RPD. In the OPD group, the jejunal limb was pulled upward for the gastrojejunostomy, and the final position of the gastrojejunostomy was antecolic, antiperistaltic, and supramesocolic.

A NGT was routinely placed during the operation for all patients. In the RPD(-)NGT group, the NGT was removed in the operating room immediately after intratracheal extubation, whereas in the RPD(+)NGT and OPD(+)NGT groups, the NGT was left in place for gastric decompression until the amount of NGT drainage was less than 300 cc, after which an oral diet was resumed upon removal of the NGT, usually on postoperative day (POD) 3 to 5. Oral intake of a clear liquid diet was allowed for RPD(-)NGT patients on postoperative day (POD) 1, followed by a regular soft diet on POD 3 whenever possible. Abdominal radiography after oral contrast intake was performed in patients with suspected DGE. Intravenous proton pump inhibitors were administered to all the patients after surgery, whereas prophylactic octreotide or prokinetic drugs were not routinely used. The indications for NGT reinsertion included persistent nausea, emesis, or clinically detected gastric distention.

Definitions of surgical complications

DGE was referred to as a clinically significant grade B or C, based on the criteria proposed by the International Study Group of Pancreatic Surgery (ISGPS) [16]. POPF was clinically relevant grade B or C pancreatic leakage, based on the 2016 new grading system by the International Study Group for Pancreatic Fistula (ISPGF) [17]. Postpancreatectomy hemorrhage (PPH) and chyle leak were classified using the standardized criteria proposed by the ISGPS [18, 19]. The radicality of resection was categorized into three groups based on the resection margin status: R0, indicating a resection without any visible or microscopic evidence of cancer at the resection margin, with a margin > 1 mm; R1, indicating a resection with no visible evidence but microscopic presence of cancer at the resection margin, with a margin ≤ 1 mm; and R2, indicating a resection with visible evidence of cancer at the resection margin. Surgical mortality was defined as death within 90 days after surgery, including the period of admission for the operation or hospital readmission.

Statistical analysis

Statistical analyses were performed using the Statistical Product and Service Solutions 26.0 version (IBM Corp., Armonk, NY, USA). Categorical variables are presented as numbers (percentages) and compared using Pearson’s χ² test or Fisher’s exact test contingency tables. All continuous data are presented as median (range) and mean ± standard deviation. One-way analysis of variance (ANOVA) was used to determine whether there were any statistically significant differences between the means of three or more independent (unrelated) groups. Student’s t-test was used to compare the means of two groups. When continuous variables did not follow a normal distribution, the Wilcoxon rank-sum test was used. A multivariate analysis was performed using binary logistic regression. Confidence intervals were set at 95%, and a p value of < 0.05 was considered statistically significant.

In total, 300 patients were included in this study after propensity score matching (PSM) and divided into three groups: RPD(-)NGT, RPD(+)NGT, and OPD(+)NGT. Each arm comprised 100 patients. The demographic characteristics of patients who underwent pancreaticoduodenectomy with or without a NG tube are presented in Table 1. The RPD(-)NGT group had a higher number of patients experiencing nausea/vomiting before operation, whereas the RPD(-)NGT group had a higher proportion of patients with an ASA physical status classification ≥ 3. However, there were no significant differences in sex, age, DM, BMI, body weight loss, malignancy, or periampullary lesions among the three groups.

Table 1

Demographic characteristics for patients undergoing pancreaticoduodenectomy without or with NG tube palcement
	RPD without NG tube	RPD with NG tube	OPD with NG tube	p value
Patients, n (%)	100	100	100
Sex				0.223
Female	41 (41%)	45 (45%)	53 (53%)
Male	59 (59%)	55 (55%)	47 (47%)
Age, year old				0.769
Median (range)	68 (13–94)	66 (29–88)	66 (23–87)
Mean ± SD	65.9 ± 13.2	65.3 ± 15.1	64.6 ± 13.0
DM	39 (39%)	30 (30%)	32 (32%)	0.386
Jaundice	45 (45%)	55 (55%)	58 (58%)	0.209
Nausea/vomiting before operation	36 (36%)	23 (23%)	19 (19%)	0.016
BMI
Median (range)	23.5 (15.4– 33.7)	23.6 (16.7–34.5)	22.6 (15.2–34.2)	0.071
Mean ± SD	23.8 ± 3.3	24.0 ± 3.7	22.9 ± 3.5
ASA physical status classification				0.003
< 3	55 (55%)	76 (76%)	72 (72%)
≥ 3	45(45%)	24 (24%)	28 (28%)
Body weight loss	38 (38%)	33 (33%)	37 (37%)	0.738
Tumor size, cm				0.386
Median (range)	3.0 (0.5–6.5)	3.0 (1.0–11.0)	3.0 (0.8–14.0)
Mean ± SD	3.0 ± 1.3	3.3 ± 1.6	3.4 ± 2.2
Malignancy	90 (90%)	91 (91%)	92 (92%)	0.885
Periampullary lesions				0.116
Pancreatic head adenocarcinoma	36 (36%)	35 (35%)	49 (49%)
Ampullary adenocarcinoma	22 (22%)	34 (35%)	22 (22%)
Distal CBD adenocarcinoma	7 (7%)	4 (4%)	1 (1%)
Duodenal adenocarcinoma	12 (12%)	8 (8%)	8 (8%)
Other malignancy	13 (14%)	12 (12%)	12 (12%)
Other benign lesions	10 (10%)	7 (7%)	5 (5%)
Chronic pancreatitis	0	2 (2%)	3 (3%)
RPD: robotic pancreaticoduodenectomy; NG: nasogastric; OPD: open pancreaticoduodenectomy; SD: standard deviation; DM: diabetes mellitus; BMI: body mass index; ASA: American Society of Anesthesiologists; CBD: common bile duct

The RPD(-)NGT group had a longer operation time compared to both the RPD(+)NGT and OPD(+)NGT groups, with a median of 10.6 vs. 7.8 vs. 7.0 hours, respectively (p < 0.001). Additionally, mesopancreas dissection level III was more frequently performed in the RPD(-)NGT group (26% vs. 15% vs. 20%, p < 0.001). However, the RPD groups had less intraoperative blood loss, with a median of 155 vs. 200 vs. 500 cc (p < 0.001), and a higher lymph node yield, with a median of 18 vs. 15 vs. 12 cc (p < 0.001). The RPD(-)NGT group also showed lower rates of vascular resection, perineural invasion, and stage I and II disease (Table 2). The PPPD, and radicality rates were similar among the three groups.

Table 2

Surgical and pathological parameters for patients undergoing pancreaticoduodenectomy without or with NG tube palcement
	RPD without NG tube	RPD with NG tube	OPD with NG tube	p value
Patients, n	100	100	100
Operation time, hour				< 0.001
Median (range)	10.6 (5.8–16.0)	7.8 (3.9–16.3)	7.0 (4.0–12.5)
Mean ± SD	10.7 ± 1.7	8.1 ± 2.4	7.5 ± 1.9
Mesopancreas dissection				< 0.001
Level 1	22 (22%)	81 (81%)	78 (78%)
Level 1I	52 (52%)	4 (4%)	2 (2%)
Level 1II	26 (26%)	15 (15%)	20 (20%)
Blood loss, c.c.				< 0.001
Median (range)	155 (0–730)	200 (1–1200)	500 (50–2750)
Mean ± SD	176 ± 136	229 ± 225	610 ± 454
Lymph node yield				< 0.001
Median (range)	18 (5–34)	15 (3–43)	12 (5–30)
Mean ± SD	18 ± 6	16 ± 7	12 ± 7
Vascular resection	5 (5%)	13 (13%)	20 (20%)	0.006
PPPD	45 (45%)	47 (47%)	51 (51%)	0.688
Radicality				0.161
R0	91 (91%)	94 (94%)	94 (94%)
R1	7 (7%)	1 (1%)	3 (3%)
R2	2 (2%)	5 (5%)	3 (3%)
Lymph node involvement	45 (45%)	45 (45%)	48 (48%)	0.886
Perineural invasion	46 (46%)	62 (62%)	62 (62%)	0.031
Lymphovascular invasion	45 (45%)	53 (53%)	28 (28%)	0.001
Stage				0.007
I and II	53 (67.9%)	73 (82.0%)	76 (87.4%)
III and IV	25 (32.1%)	16 (18.0%)	11 (12.6%)
RPD: robotic pancreaticoduodenectomy; NG: nasogastric; OPD: open pancreaticoduodenectomy; PPPD: pylorus-preserving pancreaticoduodenectomy; SD: standard deviation; R0: curative resection without residual cancer > 1 mm margin; R1: microscopic residual cancer ≤ 1 mm margin; R2: gross residual cancer

The surgical outcomes were comparable among the RPD(-)NGT, RPD(+)NGT, and OPD(+)NGT groups, except for DGE and LOS (Table 3). The occurrence of DGE was significantly lower in the RPD groups, both with and without NGT, compared to the OPD(+)NGT group (3% vs. 4% vs. 18%, p < 0.001). The LOS was also shorter in the RPD(-)NGT group, with a median of 17 days, compared to the other two groups with NGT: 22 days in the RPD(+)NGT group and 26 days in the OPD(+)NGT group (p < 0.001).

Table 3

Surgical outcomes for patients undergoing pancreaticoduodenectomy without or with NG tube palcement
	RPD without NG tube	RPD with NG tube	OPD with NG tube	p value
Patients, n	100	100	100
Surgical mortality	1 (1%)	0	2 (2%)	0.364
Surgical morbidity	58 (58%)	47 (47%)	53 (53%)	0.296
Postoperative complications				0.084
Clavien–Dindo 0	42 (42%)	53 (53%)	47 (47%)
Clavien–Dindo I	35 (35%)	35 (35%)	44 (44%)
Clavien–Dindo II	11 (11%)	3 (3%)	6 (6%)
Clavien–Dindo III	11 (11%)	8 (8%)	2 (2%)
Clavien–Dindo IV	0	1 (1%)	0
Clavien–Dindo V (death)	1 (1%)	0	1 (1%)
DGE	3 (3%)	4 (4%)	18 (18%)	< 0.001
POPF	8 (8%)	5 (5%)	8 (8%)	0.631
Intraabdominal abscess	8 (8%)	7 (7%)	2 (2%)	0.145
PPH	5 (5%)	2 (2%)	3 (3%)	0.485
Chyle leakage	20 (20%)	9 (9%)	18 (18%)	0.074
Bile leakage	0	2 (2%)	1 (1%)	0.364
Wound infection	3 (3%)	4 (4%)	4 (4%)	0.910
LOS, day				< 0.001
Median (range)	17 (6–46)	22 (6–77)	26 (10–99)
Mean ± SD	19 + 8	25 ± 15	28 ± 14
RPD: robotic pancreaticoduodenectomy; NG: nasogastric; OPD: open pancreaticoduodenectomy; DGE: delayed gastric emptying; POPF: postoperative pancreatic fistula; PPH: postpancreatectomy hemorrhage; SD: standard deviation; LOS: length of stay

Univariate analysis revealed that NG tube placement and OPD operation type were associated with an increased risk of DGE (Table 4). However, after performing multivariate analysis using binary logistic regression, only OPD operation type remained an independent risk factor for predicting DGE (Fig. 1).

Table 4

Delayed gastric emptying in patients undergoing pancreaticoduodenectomy
	DGE (+)	DGE (-)	p value
Patients, n	25	275
Placement of NG tube			0.018
(+), n = 200	22 (11.0%)	178 (98.0%)
(-), n = 100	3 (3.0%)	97 (97.0%)
Age, years old			0.549
< 65, n = 127	12 (9.4%)	115 (90.6%)
≥ 65, n = 173	13 (7.5%)	160 (92.5%)
Jaundice			0.754
(+), n = 159	14 (8.8%)	145 (91.2%)	0.084
(-), n = 141	11 (7.8%)	130 (92.2%)
DM			0.484
(+), n = 101	10 (9.9%)	91 (90.1%)
(-), n = 199	15 (7.5%)	184 (92.5%)
Malignancy			0.101
(+), n = 273	25 (9.2%)	248 (90.8%)
(-), n = 27	0	27 (100%)
LN involvement			0.530
(+), n = 138	13 (9.4%)	125 (90.6%)
(-), n = 162	12 (7.4%)	150 (92.6%)
Operation type			< 0.001
RPD, n = 200	7 (3.5%)	193 (96.5%)
OPD, n = 100	18 (18.0%)	82 (82.0%)
PPPD			0.972
(+), n = 143	12 (8.4%)	131 (91.6%)
(-), n = 157	13 (8.3%)	144 (91.7%)
DGE: delayed gastric emptying; NG: nasogastric; DM: diabetes mellitus; RPD: robotic pancreaticoduodenectomy; OPD: open pancreaticoduodenectomy; PPPD: pylorus-preserving pancreaticoduodenectomy

DGE is a common and frustrating complication that may occur after pancreaticoduodenectomy, although it is typically not life threatening. In most cases, DGE resolves spontaneously either with or without prokinetics, but it often requires several weeks or more of conservative management through NGT drainage [20]. However, DGE has a significant impact on the patient’s quality of life, prolongs the hospital stay, and increases hospital costs [11]. The incidence of DGE following pancreaticoduodenectomy is quite high, ranging from 4.5–56.1% [5–8], leading the ISGPS to propose a consensual definition of DGE in 2007 that is composed of a three-grade classification [16]. The exact pathogenesis of DGE is poorly understood and believed to involve multiple factors [7, 8, 14, 21–23]. Avoidance of stomach distention, postoperative nausea, vomiting, aspiration pneumonia, and wound dehiscence are some of the rationales behind the practice of routine placement of NGT following pancreaticoduodenectomy. Therefore, foregoing a NGT following pancreaticoduodenectomy has not been widely accepted by most pancreatic surgeons [11]. This reluctance could stem from the limited data available on this specific indication, the type of pancreatic anastomosis performed, and, notably, the high incidence of DGE associated with pancreaticoduodenectomy [5–8]. Nevertheless, the 2013 Enhanced Recovery After Surgery (ERAS®) guidelines strongly advised against preemptive use of nasogastric tubes postoperatively, as they did not improve outcomes and might impede recovery [24].

Given the high incidence of DGE after OPD, with rates ranging from 12–16%, as evidenced in our previous studies [4, 5, 15], NGT placement had been part of our routine practice following pancreaticoduodenectomy, including cases of both RPD and OPD, until March 2023. However, considering our observation of a low incidence of DGE in RPD, ranging from 3.4–4.4% according to our previous studies [4, 5], we decided to discontinue routine placement of NGT following RPD from March 2022 onwards. It is important to note that the NGT is now consistently removed in the operating room immediately after intratracheal extubation following RPD. In our previous study [5], we proposed three hypotheses to explain the lower incidence of DGE in patients who have undergone RPD. First, “food flow by gravity,” a relatively “vertical and straight” stomach position after extracorporeal hand-sewn gastrojejunostomy via a small umbilical wound might facilitate food passage downward. Second, “separation of inflammation,” “inframesocolic, antecolic, and antiperistaltic (left-sided)” gastrojejunal anastomosis could keep the stomach away from the inflamed area above the mesocolon and transverse colon. Third, “less inflammation/adhesion” with a “smaller wound and less trauma” could result in less inflammation/adhesion.

In this PSM comparative study, we observed a significantly lower incidence of DGE in the RPD group, regardless of whether the patients had a NGT. The DGE rates were 3% in the RPD(-)NGT and 4% in RPD(+)NGT groups, whereas the OPD(+)NGT group had a higher DGE rate of 18%. We found that the LOS was also shorter in the RPD(-)NGT group, with a median duration of 17 days, compared to that in the other two groups: RPD(+) NGT (22 days) and OPD(+)NGT (26 days). After conducting a multivariate analysis, we identified that only the operation type, specifically OPD, remained an independent risk factor for predicting the occurrence of DGE. The findings of this study are consistent with the conclusions of Kunstman et al. [10] who compared routine and selective NGT gastric decompression. Their study demonstrated that patients in the selective group had a decreased incidence of DGE, LOS, and time to dietary tolerance. They concluded that routine postoperative nasogastric decompression in patients undergoing pancreaticoduodenectomy was unnecessary in many cases and could adversely affect the postoperative course. Another retrospective study by Gaignard, et al. [11] also supported these findings, suggesting that absence of systematic nasogastric decompression after pancreaticoduodenectomy might reduce postoperative complications, DGE, and LOS [11]. Based on the results of our current study, we propose that foregoing the placement of a NGT is a feasible approach in RPD cases.

Surgical outcomes, including mortality, Clavien–Dindo complication grades, and POPF, were comparable between the RPD(-)NGT, RPD(+)NGT, and RPD(+)NGT groups in this study. A retrospective study by Choi et al. [1] concluded that the routine insertion of a NGT in patients undergoing pancreaticoduodenectomy provided no advantages in terms of postoperative complications and could potentially increase the occurrence of postoperative pulmonary complications. Additionally, NGT insertion could cause discomfort to patients after surgery. Miyazawa et al. [3] suggested that no NGT management after pancreaticoduodenectomy could actually enhance the quality of life by reducing discomfort and allowing advancement to a fast-track program. In a randomized clinical trial by Bergeat et al. [25] studying nasogastric decompression vs. no decompression after pancreaticoduodenectomy, no significant difference was observed in the occurrence of postoperative complications classified as Clavien-Dindo classification grade II or higher between systematic NGT decompression and no decompression, indicating that avoiding systematic nasogastric decompression is safe. Kleive et al. [12] conducted a prospective observational study and concluded that routine use of NGT after pancreaticoduodenectomy was not justified within an ERAS setting. They demonstrated that immediate removal of the NGT after the procedure could be performed safely, and reinsertion on demand was rarely necessary during uncomplicated courses. Based on these findings, this study further supports the notion that foregoing a NGT has no negative impact on the safety of RPD cases.

This study was limited by the retrospective identification of the variables. Therefore, selection bias is inevitable despite attempts to mitigate this by using propensity score matching to mimic some characteristics of a randomized controlled trial.

The RPD without NGT group exhibited lower DGE and shorter LOS. Only the operation type, specifically OPD, remained an independent risk factor for predicting the occurrence of DGE after multivariate analysis. Surgical outcomes were comparable among the RPD(-)NGT, RPD(+)NGT, and RPD(+)NGT groups. Based on the findings of our current study, it is feasible to forego a NGT in RPD cases without negative impact on safety.

Author contributions

The authors confirm contribution to the paper as follows: study conception and design: B.-U.S., S.-E.W., B.-S.S., S.-C.C., and Y.-M.S.; data collection: B.-U.S., S.-E.W., B.-S.S., Y.-M.S.; analysis and interpretation of results: B.-S.S., and B.-U.S.; draft manuscript preparation: B.-U.S., S.-E.W., B.-S.S., S.-C.C.,. All authors reviewed the results and approved the final version of the manuscript.

Funding

This work was supported by grants from Taipei Veterans General Hospital (V112C-009, V112C-188, and V112B-001), the Ministry of Science and Technology (NSTC 112-2314-B-075-016 -), and the Ministry of Health and Welfare (MOHW111-TDU-B-221-014015).

Data availability

The data that support the findings of this study are available from the corresponding author, Bor-Uei Shyr, upon reasonable request.

Declarations Conflict of interest

The author(s), Bor-Uei Shyr, Shin-E Wang, Bor-Shiuan Shyr, Shih-Chin Chen, and Yi-Ming Shyr have no conflicts of interest or financial ties to disclose.

Ethics approval

The study adhered to the tenets of the Declaration of Helsinki, and this study adhered to all local regulatory requirements applicable to clinical investigations. This study was approved by the Institutional Review Board (IRB) of Taipei Veterans General Hospital (IRB-TPEVGH No. 2023-07-007CC).

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Consent to publish

All patients were given complete information on the risks and benefits of the procedure and gave their written consent.

Acknowledgments

The authors would like to acknowledge the support of the Biobank of Taipei Veterans General Hospital, Common Well Foundation, and statistical team of Taipei Veterans General Hospital.

Choi YY, Kim J, Seo D, Choi D, Kim MJ, Kim JH, Lee KJ, Hur KY (2011) Is routine nasogastric tube insertion necessary in pancreaticoduodenectomy? J Korean Surg Soc 81:257–262.
Fisher WE, Hodges SE, Cruz G, Artinyan A, Silberfein EJ, Ahern CH, Jo E, Brunicardi FC (2011) Routine nasogastric suction may be unnecessary after a pancreatic resection. HPB (Oxford) 13:792–796.
Miyazawa M, Kawai M, Hirono S, Okada KI, Kitahata Y, Kobayashi R, Ueno M, Hayami S, Miyamoto A, Yamaue H (2011) Previous upper abdominal surgery is a risk factor for nasogastric tube reinsertion after pancreaticoduodenectomy. Surgery 170:1223–1230.
Wang SE, Shyr BU, Chen SC, Shyr YM (2018) Comparison between robotic and open pancreaticoduodenectomy with modified Blumgart pancreaticojejunostomy: A propensity score-matched study. Surger 164:1162–1167.
Mao SH, Shyr BS, Chen SC, Wang SE, Shyr YM, Shyr BU (2022) Risk factors for delayed gastric emptying in pancreaticoduodenectomy. Sci Rep 12:22270.
Mack LA, Kaklamanos IG, Livingstone AS, Levi JU, Robinson C, Sleeman D, Franceschi D, Bathe OF (2004) Gastric decompression and enteral feeding through a double-lumen gastrojejunostomy tube improves outcomes after pancreaticoduodenectomy. Ann Surg 240:845–851.
Snyder RA, Ewing JA, Parikh AA (2020) Delayed gastric emptying after pancreaticoduodenectomy: A study of the national surgical quality improvement program. Pancreatology 20:205–210.
Lee YH, Hur YH, Kim HJ, Kim CY, Kim JW (2021) Is delayed gastric emptying associated with pylorus ring preservation in patients undergoing pancreaticoduodenectomy? Asian J Surg 44:137–142.
Park JS, Kim JY, Kim JK, Yoon DS (2016) Should Gastric Decompression be a Routine Procedure in Patients Who Undergo Pylorus-Preserving Pancreatoduodenectomy? World J Surg 40:2766–2770.
Kunstman JW, Klemen ND, Fonseca AL, Araya DL, Salem RR (2013) Nasogastric Drainage May Be Unnecessary after Pancreaticoduodenectomy: A Comparison of Routine vs Selective Decompression. J Am Coll Surg 2013, 217:481–488.
Gaignard E, Bergeat D, Courtin-Tanguy L, Rayar M, Merdrignac A, Robin F, Boudjema K, Beloeil H, Meunier B, Sulpice L (2018) Is systematic nasogastric decompression after pancreaticoduodenectomy really necessary? Langenbecks Arch Surg 403:573–580.
Kleive D, Sahakyan MA, Labori KJ, Lassen K (2019) Nasogastric Tube on Demand is Rarely Necessary After Pancreatoduodenectomy Within an Enhanced Recovery Pathway. World J Surg 43:2616–2622.
Flick KF, Soufi M, Yip-Schneider MT, Simpson RE, Colgate CL, Nguyen TK, Ceppa EP, House MG, Zyromski NJ, Nakeeb A, Schmidt CM (2021) Routine Gastric Decompression after Pancreatoduodenectomy: Treating the Surgeon? J Gastrointest Surg 25:2902–2907.
Jung JP, Zenati MS, Dhir M, Zureikat AH, Zeh HJ, Simmons RL, Hogg ME (2018) Use of Video Review to Investigate Technical Factors That May Be Associated With Delayed Gastric Emptying After Pancreaticoduodenectomy. JAMA Surg 153:918–927.
Wang SE, Chen SC, Shyr BU, Shyr YM (2016 Comparison of Modified Blumgart pancreaticojejunostomy and pancreaticogastrostomy after pancreaticoduodenectomy. HPB (Oxford) 18:229–235.
Wente MN, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, Neoptolemos JP, Padbury RT, Sarr MG, Traverso LW, Yeo CJ, Büchler MW (2007) Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 142:761–768.
Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, Allen P, Andersson R, Asbun HJ, Besselink MG, Conlon K, Del Chiaro M, Falconi M, Fernandez-Cruz L, Fernandez-Del Castillo C, Fingerhut A, Friess H, Gouma DJ, Hackert T, Izbicki J, Lillemoe KD, Neoptolemos JP, Olah A, Schulick R, Shrikhande SV, Takada T, Takaori K, Traverso W, Vollmer CR, Wolfgang CL, Yeo CJ, Salvia R, Buchler M (2017) The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 Years After. Surgery 161:584–591.
Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, Neoptolemos JP, Padbury RT, Sarr MG, Yeo CJ, Büchler MW (2007) Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 142:20–25.
Besselink MG, van Rijssen LB, Bassi C, Dervenis C, Montorsi M, Adham M, Asbun HJ, Bockhorn M, Strobel O, Büchler MW, Busch OR, Charnley RM, Conlon KC, Fernández-Cruz L, Fingerhut A, Friess H, Izbicki JR, Lillemoe KD, Neoptolemos JP, Sarr MG, Shrikhande SV, Sitarz R, Vollmer CM, Yeo CJ, Hartwig W, Wolfgang CL, Gouma DJ; International Study Group on Pancreatic Surgery (2017) Definition and classification of chyle leak after pancreatic operation: A consensus statement by the International Study Group on Pancreatic Surgery. Surgery 161:365–372.
Miyazaki Y, Oda T, Shimomura O, Enomoto T, Akashi Y, Hisakura K, Ohara Y, Kimura S, Hashimoto S, Ohkohchi N (2019) Retrocolic Gastrojejunostomy After Pancreaticoduodenectomy: A Satisfactory Delayed Gastric-Emptying Rate. Pancreas 48:579–584.
Ellis RJ, Gupta AR, Hewitt DB, Merkow RP, Cohen ME, Ko CY, Bilimoria KY, Bentrem DJ, Yang AD (2019) Risk factors for post-pancreaticoduodenectomy delayed gastric emptying in the absence of pancreatic fistula or intra-abdominal infection. J Surg Oncol 119:925–931.
Panwar R, Pal S (2017) The International Study Group of Pancreatic Surgery definition of delayed gastric emptying and the effects of various surgical modifications on the occurrence of delayed gastric emptying after pancreatoduodenectomy. Hepatobiliary Pancreat Dis Int 16:353–363.
Eisenberg JD, Rosato EL, Lavu H, Yeo CJ, Winter JM (2015) Delayed Gastric Emptying After Pancreaticoduodenectomy: an Analysis of Risk Factors and Cost. J Gastrointest Surg 19:1572–1580.
Lassen K, Coolsen MM, Slim K, Carli F, de Aguilar-Nascimento JE, Schäfer M, Parks RW, Fearon KC, Lobo DN, Demartines N, Braga M, Ljungqvist O, Dejong CH; Enhanced Recovery After Surgery (ERAS) Society, for Perioperative Care; European Society for Clinical Nutrition and Metabolism (ESPEN) (2013) Guidelines for perioperative care for pancreaticoduodenectomy: Enhanced Recovery After Surgery (ERAS®) Society recommendations. World J Surg 37:240–258.
Bergeat D, Merdrignac A, Robin F, Gaignard E, Rayar M, Meunier B, Beloeil H, Boudjema K, Laviolle B, Sulpice L (2020) Nasogastric Decompression vs No Decompression After Pancreaticoduodenectomy: The Randomized Clinical IPOD Trial. JAMA Surg 155:e202291.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Impact of foregoing a nasogastric tube in robotic pancreaticoduodenectomy: a propensity score-matched study

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Study endpoints

Propensity score matching (PSM) strategy

Surgical technique

Definitions of surgical complications

Statistical analysis

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1