Influence of Body Mass Index on Perioperative Outcomes Following Pancreaticoduodenectomy

Abstract

Background:

Overweight and obesity are increasing year by year all over the world, and there is a correlation between overweight and obesity and the risk of pancreatic cancer. However, the relationship between overweight and obesity and perioperative outcomes of pancreaticoduodenectomy (PD) was controversial. The purpose of this study was to investigate the effect of body mass index (BMI) on the perioperative outcome of PD.

Methods:

This study retrospectively evaluated 227 patients who underwent PD from 2015 to 2019. The patients were divided into three groups: underweight group (BMI <18.5 kg/m²), normal weight group (18.5 ≤ BMI <25 kg kg/m²), and overweight group (BMII ≥25 kg/m²). The association between different BMI groups and different perioperative results was discussed. Finally, the independent risk factors of clinically relevant-postoperative pancreatic fistula (CR-POPF) were analyzed by multivariate logistic regression.

Results:

The level of preoperative albumin was higher in patients of overweight group (P = .03). The incidence of hypertension increased gradually in the three BMI groups (P = . 039). The preoperative median CA19–9 level was significantly higher in the underweight group than that in the control groups (P = .001). The median operation time in the high BMI group was significantly longer than that in the other two groups. High BMI was an independent risk factor influencing CR-POPF after PD (P = .022, odds ratio 2.253, 95% confidence interval 1.123–4.518).

Conclusions:

Operation time of PD was increased in patients with high BMI. High BMI was an independent risk factor for the incidence of CR-POPF after PD. However, PD surgery is safe and feasible for patients with different BMI, and overweight and obese patients should not refuse PD surgery because of their BMI.

Introduction

Pancreaticoduodenectomy (PD) is one of the most complex operations in general surgery, and is the first choice for the treatment of various benign and malignant diseases around the head of the pancreas, the lower segment of the common bile duct and the ampulla of the duodenum.^1–3 With the improvement of surgical techniques and perioperative management, the hospital mortality after PD in high-volume pancreatic centers had been reduced to <4%, but the postoperative morbidity remains high, ranging from 39% to 60%.^4–10 There were many factors that affect the postoperative morbidity of PD, such as gender, age, body mass index (BMI), and surgical methods,^11–13 and BMI was one of the most common risk factors for postoperative complications. For the past four decades, the rate of overweight and obesity among adults had risen from 21% to 40% around the world.¹⁴

Epidemiological studies had shown that overweight and obesity are not only associated with mortality and morbidity of cardiovascular and cerebrovascular diseases, type 2 diabetes, chronic kidney disease, multiple cancers, and musculoskeletal diseases,¹⁵ but also important risk factors for surgical outcomes and prognosis of many diseases.^16–18 However, controversy remained as to the effects of high BMI on perioperative outcomes of PD. For instance, Gaujoux et al.¹⁹ reported that high BMI was related to the increase of intraoperative blood loss in PD, on the contrary, Su et al.²⁰ thought that BMI was not a risk factor for the increased intraoperative blood loss. Zarzavadjian et al.²¹ reported that higher BMI was an important predictor of POPF Grade B/C after PD, whereas Shimizu et al.²² presented a completely opposite point of view that there was no significant difference in the incidence of postoperative POPF Grade B/C between overweight and obese patients who underwent PD compared with normal weight patients. In view of these contradictory results, our aim was to elucidate the effect of BMI on the perioperative outcomes of PD.

Materials and Methods

A retrospective analysis was performed on 232 Chinese patients who underwent PD in the Third Affiliated Hospital of Soochow University from January 2015 to March 2019. The types of diseases included pancreatic ductal adenocarcinoma (n = 87), ampullary adenocarcinoma (n = 11), lower cholangiocarcinoma (n = 45), duodenal adenocarcinoma (n = 35), intraductal papillary mucinous neoplasm (n = 22), pancreatic neuroendocrine tumor (n = 3), chronic pancreatitis (n = 4), and other types of tumor (n = 20). Surgical methods included open pancreaticoduodenectomy (OPD; n = 89) and laparoscopic pancreaticoduodenectomy (LPD; n = 138) of which 10 cases underwent portal vein or superior mesenteric vein resection and reconstruction. This study excluded 5 patients who underwent LPD combined with transverse colectomy (n = 3), combined adrenalectomy (n = 1), and lack of BMI data (n = 1). Finally, a total of 227 patients were enrolled in the study. According to the guidelines of the World Health Organization, BMI was calculated as weight (kg) divided by the square of height (m²). The enrolled patients were divided into three groups: underweight group (BMI <18.5 kg/m²), normal weight group (18.5 ≤ BMI <25 kg/m²), and overweight group (BMI ≥25 kg/m²).

The collected data included (1) preoperative clinical data: gender, age, BMI, initial symptom, concomitant disease (hypertension/diabetes), total bilirubin, serum albumin, tumor markers (CA19–9, CEA), preoperative reduction of jaundice (yes/no), American Society of Anesthesiologists classification (ASA) rating; (2) operation-related index: surgical method (OPD/LPD), operation time, estimated blood loss, vascular reconstruction, pancreatic texture (soft/hard); (3) postoperative related indicators: time to oral intake, removal time of all drainage tubes, postoperative complications (postoperative pancreatic fistula [POPF], postpancreatectomy hemorrhage [PPH], delayed gastric emptying [DGE]), Clavien–Dindo classification, length of stay (LOS), hospitalization expenses, and in-hospital mortality. (4) Histopathological features: pathological type, tumor size (<2 cm or ≥2 cm), histological diagnosis (malignant/benign), harvested lymph nodes, lymph node status (positive/negative), R0 resection (yes/no), and peripheral nerve invasion (yes/no). This study was approved by the Ethics Committee of the Third Affiliated Hospital of Soochow University, and was conducted in accordance with its relevant guidelines and regulation. Informed consent was obtained from all participants.

Surgical procedure

All patients received standard PD, whether it was OPD or LPD. All patients were reconstructed with Child's method (pancreaticojejunostomy, choledochojejunostomy, and gastrojejunostomy),²³ and the internal supporting tube was routinely placed in the pancreatic duct. Preventive closed drainage tubes were placed around the pancreaticojejunum and choledochojejunostomy, respectively. All PDs were performed by the same experienced surgeon (C.X.M.).

Definitions

According to the standard of the International Study Group of Pancreatic Fistula (ISGPS), POPF was defined as any measurable drainage fluid on or after the third day postoperative with amylase level higher than three times the upper limit of normal serum amylase in each particular institution.²⁴ According to the severity, POPF was further classified as biochemical leak (BL) and POPF Grade B/C. BL was defined as biochemical fistula with no significant clinical effect, and POPF Grade B/C were defined as clinically related POPF (CR-POPF). PPH after PD was defined by the onset time, location, and severity.²⁵ DGE was defined as if the patient needs for maintenance of a nasogastric tube (NGT) for >3 days or reinserting the NGT after postoperative day (POD) 3 due to persistent nausea or vomiting and cannot resume a standard diet by POD 7.²⁶ Clavien–Dindo classification was based on the standard proposed by Clavien and Dindo.²⁷

Statistical analysis

SPSS version 20.0 (SPSS, Chicago, IL) was used for statistical analysis. According to the data distribution, continuous variables are expressed as mean ± standard deviation (SD) or median (and quartile range, IQR). For continuous data, Student's t-test or Mann–Whitney U test were used to compare the differences. The classified variables were analyzed by chi-square test or Fisher's exact test. Logistic regression analysis was used to analyze the risk factors of CR-POPF. The variables of P < .20 in univariate analysis were included in multivariate logistic regression analysis. P < .05 was considered statistically significant.

Results

Demographic data and preoperative variables

A total of 227 patients were included in the study and were divided into three groups: 22 cases of underweight group (9.7%), 160 cases of normal weight group (70.5%), and 45 cases of overweight group (19.8%). There were no significant differences in gender, age, history of diabetes, preoperative reduction of jaundice, preoperative total bilirubin, CEA, and ASA score among the three groups (Table 1). Preoperative albumin among different BMI groups was significantly different (P = .030). The incidence of hypertension in underweight, normal weight, and overweight groups increased from 13.6% to 39.4% and 44.4%, respectively (P = .039). In addition, CA19–9 was significantly different in three BMI groups (P = .001).

Table 1.

Demographic Data and Preoperative Variables Among the Three Groups

Variable	Overall	BMI <18.5	18.5 ≤ BMI <25	BMI ≥25	P value
Total number	227	22 (9.7%)	160 (70.5%)	45 (19.8%)
Gender					.414
Male (n, %)	143 (63.0%)	11 (50.0%)	103 (64.4%)	29 (64.4%)
Female (n, %)	83 (37.0%)	11 (50.0%)	57 (35.6%)	16 (35.6%)
Age (years ± SD)	66.3 ± 9.0	66.9 ± 9.5	66.7 ± 8.8	64.5 ± 9.3	.333
Hypertension (n, %)	86 (37.9%)	3 (13.6%)	63 (39.4%)	20 (44.4%)	.039
Diabetes (n, %)	43 (18.9%)	3 (13.6%)	33 (20.6%)	7 (15.6%)	.596
Preoperative albumin (g/dL ± SD)	36 ± 6.1	34.1 ± 8.3	35.7 ± 5.7	38 ± 6.1	.030
Preoperative bilirubin (μM, IQR)	24.9 (9.8–118.9)	32 (12.1–106.1)	23.5 (9.6–118.4)	29.3 (8.6–142.4)	.934
Preoperative biliary drainage (n, %)	22 (9.7%)	2 (9.1%)	16 (10.0%)	4 (8.9%)	.97
Preoperative CA19–9 (U/L, IQR)	67.7 (17.6–179.5)	158.5 (25.3–448.1)	66 (17.6–159.9)	65 (15.4–314.4)	.001
Preoperative CEA (μg/L, IQR)	3.1 (1.9–4.5)	3.3 (2.4–5.8)	3.2 (1.9–4.6)	2.6 (1.5–3.5)	.880
ASA score					.144
I-II (n, %)	172 (75.8%)	17 (77.3%)	116 (72.5%)	39 (86.7%)
III-IV (n, %)	55 (24.2%)	5 (22.7%)	44 (27.5%)	6 (13.3%)

P < .05 was considered statistically significant.

ASA, American Society of Anesthesiologists classification; BMI, body mass index; IQR, interquartile range; SD, standard deviation.

The influence of BMI on intraoperative findings

Table 2 described different surgical variables among the three groups (operation method [OPD/LPD], operation time, intraoperative blood loss, pancreatic texture [soft/hard], vascular reconstruction [yes/no]). There was no statistically significant difference in pancreatic texture among the three BMI groups (P = .204). The operation time of the overweight group was significantly longer than that of the other two groups (P = .004). Surprisingly, there was no significant difference in intraoperative blood loss between different BMI groups (P = .113). Although the laparoscopic rate was higher in the overweight group, there was no significant difference between the three groups (P = .112), and there was no significant correlation between the vascular reconstruction and BMI (P = .130).

Table 2.

Comparison of Intraoperative Findings in the Three Groups

Variable	Overall (n = 227)	BMI <18.5 (n = 22)	18.5 ≤ BMI <25 (n = 160)	BMI ≥25 (n = 45)	P value
Surgical method					.112
OPD (n, %)	89 (39.2%)	13 (59.1%)	61 (38.1%)	15 (33.3%)
LPD (n, %)	138 (60.8%)	9 (40.9%)	99 (61.9%)	30 (66.7%)
Pancreatic texture					.210
Soft (n, %)	194 (85.5%)	16 (72.7%)	139 (86.9%)	39 (86.7%)
Hard (n, %)	33 (15.5%)	6 (27.3%)	21 (13.1%)	6 (13.3%)
Operative time (min, IQR)	340 (280–415)	315 (251–430)	325 (271–390)	385 (300–483)	.004
Estimated blood loss (mL, IQR)	200 (150–400)	375 (200–525)	200 (150–400)	300 (200–500)	.113
Vascular reconstruction (n, %)	10 (4.4%)	1 (4.5%)	9 (5.6%)	0 (0.0%)	.298

P < .05 was considered statistically significant.

BMI, body mass index; IQR, interquartile range; LPD, laparoscopic pancreaticoduodenectomy; OPD, open pancreaticoduodenectomy.

The effect of BMI on pathological results

The pathological results of different BMI groups were compared, including tumor size, histological diagnosis (malignant/benign), harvest lymph nodes, lymph node status (positive/negative), R0 resection (yes/no), and peripheral nerve invasion (yes/no) (Table 3). There was no significant correlation between mass nature and different BMI (P = .251). Similarly, the association between different BMI groups and other pathological results was not statistically significant.

Table 3.

Pathological Findings in the Three Groups

Variable	Overall (n = 227)	BMI <18.5 (n = 22)	18.5 ≤ BMI <25 (n = 160)	BMI ≥25 (n = 45)	P value
Tumor size (cm)					.497
<2 cm (n, %)	60 (26.4%)	5 (22.7%)	40 (25.0%)	15 (33.3%)
≥2 cm (n, %)	167 (73.6%)	17 (77.3%)	120 (75.0%)	30 (66.7%)
Histological diagnosis					.251
Benign (n, %)	49 (21.6%)	2 (9.1%)	35 (21.9%)	12 (26.7%)
Malignant (n, %)	178 (78.4%)	21 (90.9%)	125 (78.1%)	33 (73.3%)
Harvested lymph nodes (n, IQR)	11 (6–15)	14 (5–17)	11 (7–15)	11 (6–15)	.712
Lymph node metastasis (n, %)	72 (31.7%)	8 (36.4%)	51 (31.9%)	13 (28.9%)	.827
R0 resection (n, %)	8 (3.5%)	1 (4.5%)	5 (3.1%)	2 (4.4%)	.733
Perineural infiltration	81 (35.7%)	9 (40.9%)	58 (36.2%)	14 (31.1%)	.592
Site					.501
Pancreatic tumor	136 (59.9%)	14 (63.6%)	100 (62.5%)	22 (48.9%)
Low CBD tumor	48 (21.1%)	5 (22.7%)	32 (20.0%)	11 (24.4%)
Duodenal tumor	43 (18.9%)	3 (13.6%)	28 (17.5%)	12 (26.7%)

BMI, body mass index; CBD, Common bile duct; IQR, interquartile range.

The effect of BMI on postoperative events after PD

Postoperative events (postoperative complications [POPF, PPH, and DGE], Clavien–Dindo classification, time to oral intake, LOS, in-hospital cost, and in-hospital mortality) are shown in Table 4. The incidence of POPF increased with the rise in BMI among the three groups (P = .001). Meanwhile, the incidence of CR-POPF in the overweight group was also significantly higher than other groups (P = .029). However, there was no significant difference in PPH and DGE among different BMI groups (P = .874 and P = .799). There was also no significant difference in LOS, in-hospital cost, and in-hospital mortality among three groups.

Table 4.

Postoperative Findings in the Three Groups

Variable	Overall (n = 227)	BMI <18.5 (n = 22)	18.5 ≤ BMI <25 (n = 160)	BMI ≥25 (n = 45)	P value
POPF (n, %)	79 (34.8%)	2 (9.1%)	53 (33.1%)	24 (53.3%)	.001
CR-POPF (n, %)	35 (15.4%)	0 (0.0%)	24 (15.0%)	11 (24.4%)	.029
PPH (n, %)	13 (5.7%)	2 (9.1%)	9 (5.6%)	2 (4.4%)	.799
DGE (n, %)	22 (9.7%)	1 (4.5%)	17 (10.6%)	4 (8.9%)	.874
Clavien–Dindo classification					.444
<III (n, %)	199 (87.7%)	19 (86.4%)	138 (86.2%)	42 (93.3%)
≥III (n, %)	28 (12.3%)	3 (13.6%)	22 (13.8%)	3 (6.7%)
All drainage tube removal time (days, IQR)	11 (9–18)	10 (8–12)	11 (9–18)	14 (10–21)	.245
Time to oral intake (days, IQR)	6 (4–8)	6 (4–7)	5 (3–8)	6 (5–7)	.075
Length of stay (days, IQR)	16 (12–24)	15 (11–18)	15 (12–24)	18 (13–26)	.495
Hospitalization expenses (USD, IQR)	11564.1 (9436.2–14,975.4)	10990.2 (8464.5–16,781.9)	12229.5 (10,075.8–15,607.5)	10136.2 (8125.4–12,765.3)	.234
In-hospital mortality (n, %)	16 (7.0%)	1 (4.5%)	13 (8.1%)	2 (4.4%)	.766

BMI, body mass index; CR-POPF, clinically relevant postoperative pancreatic fistula; DGE, delayed gastric emptying; IQR, interquartile range; POPF, postoperative pancreatic fistula; PPH, postpancreatectomy hemorrhage.

Risk factors influencing the occurrence of CR-POPF

Of the 35 patients with CR-POPF, 24 patients were in the normal BMI group and 11 patients were from the high BMI group. Four variables that had a P value not >0.20 by univariate analysis were taken into multivariate analysis (BMI group, intraoperative blood loss, tumor size, and lymph node status), and only BMI was an independent risk factor for CR-POPF (P = .022, odds ratio 2.253, 95% confidence interval 1.123–4.518; Table 5).

Table 5.

Univariate and Multivariate Analysis of Risk Factors for Clinically Relevant Postoperative Pancreatic Fistula After Pancreaticoduodenectomy

Variable	Univariate analysis		Multivariate analysis
	P value	Odds ratio (95% CI)	P value	Odds ratio (95% CI)
Age (years)	.308	1.022 (0.980–1.065)	—	—
Gender (male/female)	.690	0.861 (0.412–1.799)	—	—
BMI^*	.011	2.440 (1.226–4.859)	.022	2.253 (1.123–4.518)
Hypertension (yes/no)	.510	1.278 (0.615–2.654)	—	—
Diabetes (yes/no)	.862	1.083 (0.439–2.675)	—	—
Preoperative CA19–9 (U/L)	.439	0.999 (0.998–1.001)	—	—
Preoperative CEA (μg/L)	.319	0.949 (0.857–1.052)	—	—
Preoperative albumin (g/dL)	.733	1.010 (0.952–1.072)	—	—
Preoperative bilirubin (μM)	.252	1.002 (0.999–1.006)	—	—
Preoperative biliary drainage (yes/no)	.323	1.716 (0.589–5.000)	—	—
ASA score (I-II/III-IV)	.282	1.542 (0.700–3.394)	—	—
Surgical method (OPD/LPD)	.786	1.109 (0.527–2.334)	—	—
Operative time (min)	.462	0.999 (0.995–1.002)	—	—
Estimated blood loss (mL)	.111	1.001 (1.000–1.002)	.230	1.001 (1.000–1.002)
Vascular reconstruction (yes/no)	.631	0.598 (0.073–4.875)	—	—
Pancreatic texture (soft/hard)	.572	0.725 (0.238–2.209)	—	—
Histological diagnosis (benign/malignant)	.278	0.637 (0.283–1.437)	—	—
Tumor size (cm, <2 or ≥2)	.051	0.473 (0.223–1.004)	.181	0.580 (0.261–1.288)
Lymph node metastasis (yes/no)	.011	0.488 (0.202–1.178)	.288	0.605 (0.240–1.527)
R0 resection (yes/no)	.816	0.777 (0.093–6.521)	—	—

(BMI <18.5 or 18.5 ≤ BMI <25 or BMI ≥25).

ASA, American Society of Anesthesiologists classification; BMI, body mass index; CI, confidence interval; LPD, laparoscopic pancreaticoduodenectomy; OPD, open pancreaticoduodenectomy; OR, odds ratio.

Discussion

The prevalence of overweight and obesity in the world was related to a variety of chronic diseases, including hypertension, diabetes, and chronic kidney disease.¹⁵ This was consistent with our report that hypertension was at higher risk in the high BMI group (P = .039). There was a significant relationship between preoperative albumin level and different BMI group (P = .030). BMI can indirectly reflect the preoperative nutritional status, compared with normal people, patients with high BMI had more energy reserves and stronger inflammatory response to surgical invasion.^28,29 These beneficial effects would help to reduce the incidence of postoperative complications and provide a corresponding explanation for the fact that there was no significant increase in postoperative mortality with high BMI value in our study.

With the maturity of surgical technique and the improvement of perioperative treatment and postoperative management, the in-hospital mortality after PD had been reduced to <4% in many large-capacity hospitals, but the incidence of postoperative complications was still high. Pancreatic fistula, as the main complication after PD, accounts for 14%–50% of postoperative complications.^10,30–32 Our study found that the risk of pancreatic fistula and clinical pancreatic fistula was associated with high BMI (P = .001 and P = .029), which was consistent with the result reported by El Nakeeb et al.³³ In addition, in our study, high BMI was an independent risk factor for CR-POPF (P = .022). As one of the most dangerous complications, CR-POPF needs more effective preoperative fistula predictive risk score, finer intraoperative operation, and better postoperative management to prevent and treat it.^34,35

According to previous studies, patients with high BMI had a higher incidence of soft pancreas, and patients with soft pancreas had a higher risk of pancreatic fistula after PD.³⁶ However, our study demonstrated that there was no significant difference in pancreatic texture among the three groups (P = .210), which was consistent with the report by Tsai et al. that BMI had nothing to do with pancreatic texture.³⁷ On the one hand, perhaps BMI is not the key factor to determine the texture of the pancreas; on the other hand, the texture of the pancreas is subjectively judged by the surgeon, so there may be a deviation in individual evaluation.

We did not find a significant correlation between DGE, PPH, and different BMI groups (P = .799 and P = .874), which was consistent with the results reported by Shamali and Shimizu et al.^12,22 In the past few decades, laparoscopic technology had developed rapidly, compared with open surgery; laparoscopic technology had advantages of smaller trauma, less postoperative pain, and faster postoperative recovery.³⁸ However, among the groups of BMI, we did not find that different BMI groups had statistical significance in open and minimally invasive groups (P = .112). Similarly, Poves et al. reported that different BMI groups had no statistical significance in the choice of surgical method (P = .71).³⁹

Most studies had shown a significant increase in the risk of pancreatic cancer in people with high BMI. Michaud et al.⁴⁰ believed that obesity significantly increases the risk of pancreatic cancer; Genkinger et al.⁴¹ reported that overweight people were significantly more likely to develop pancreatic cancer than people with normal weight. The aforementioned two studies had the following explanations: adipokines such as leptin, adiponectin and tumor necrosis factor α secreted by adipocytes promote the occurrence and development of tumor through various carcinogenic signal pathway.⁴²

In addition, obesity can increase hypoxia in adipose tissue, which can also promote the growth of tumor cells by activating hypoxia-induced factor-1 signal pathway.⁴³ This was contradictory to the fact that different BMI groups in our study showed no significant association in histological diagnosis (malignant/benign; P = .251). The explanation would be that pancreatic cancer was difficult to detect in the early stage, and long-term malignant tumor depleted the body, resulting in low BMI. Therefore, the incidence of pancreatic cancer in the low BMI group was relatively higher in this study.

There were some limitations in our study. The first was that no follow-up data were involved (disease-free survival and overall survival) for these 227 patients. The second limitation was that it was difficult to assess the original weight of patients with a history of weight loss, which was a common limitation in patients with pancreatic cancer. Another limitation was that obese patients (BMI ≥30 kg/m²) were included in the “overweight” cohort due to the small sample size of this study, which may confuse the perioperative outcomes of overweight patients because obese patients had poorer perioperative outcomes than overweight patients.³⁷ The last possible limitation was that the WHO expert consultation report pointed out that the existing global BMI standard was not suitable for Asian population because different race and dietary structure had a various impact on the relation between BMI and the percentage of body fat.⁴⁴ In the future, more researches are needed to determine the BMI standard applicable to the Asian population.

In conclusion, operation time of PD was increased in patients with high BMI, and high BMI was an independent risk factor for the incidence of CR-POPF after PD. However, PD surgery is safe and feasible for patients with different BMI, and overweight and obese patients should not refuse PD surgery because of their BMI.

Authors' Contribution

Y.A. took charge of conceiving and designing the study; B.X. and G.Z. were responsible for collecting the data and analyzing and interpreting the data; T.T. and Y.T. took charge of writing the article; Y.Z. was responsible for providing critical revisions; W.C. and X.C. were in charge of experimental design, financial support, and approving the final version of the article.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported by the National Natural Science Foundation of China (81602054), Applied Basic Research of Changzhou Technology Bureau (CJ20190093), Major Science and Technology Project of Changzhou Health Commission (ZD201906), and The “Six One Project” top-notch talent research project of high-level health talents of Jiangsu Provincial Health Commission (LGY2019022).

References

Yuan

, et al. Postoperative complications in elderly patients following pancreaticoduodenectomy lead to increased postoperative mortality and costs. A retrospective cohort study. Int J Surg, 2018; 60:204–209.

Williams

, et al. Impact of obesity on perioperative morbidity and mortality after pancreaticoduodenectomy. J Am Coll Surg, 2009; 208:210–217.

Hata

, et al. Effect of hospital volume on surgical outcomes after pancreaticoduodenectomy: A systematic review and meta-analysis. Ann Surg, 2016; 263:664–672.

Schneider

, et al. Provider versus patient factors impacting hospital length of stay after pancreaticoduodenectomy. Surgery, 2013; 154:152–161.

Kim

, Jo

, Kim

. Impact of surgical volume on nationwide hospital mortality after pancreaticoduodenectomy. World J Gastroenterol, 2012; 18:4175–4181.

de Wilde

, et al. Impact of nationwide centralization of pancreaticoduodenectomy on hospital mortality. Br J Surg, 2012; 99:404–410.

Yoshioka

, et al. Impact of hospital volume on hospital mortality, length of stay and total costs after pancreaticoduodenectomy. Br J Surg, 2014; 101:523–529.

Pugalenthi

, et al. Postoperative complications and overall survival after pancreaticoduodenectomy for pancreatic ductal adenocarcinoma. J Surg Oncol, 2016; 113:188–193.

De Pastena

, et al. Impact of preoperative biliary drainage on postoperative outcome after pancreaticoduodenectomy: An analysis of 1500 consecutive cases. Dig Endosc, 2018; 30:777–784.

10.

Nappo

, et al. The evolution of post-operative pancreatic fistula (POPF) classification: A single-center experience. Pancreatology, 2019; 19:449–455.

11.

Ellis

, et al. Risk factors for post-pancreaticoduodenectomy delayed gastric emptying in the absence of pancreatic fistula or intra-abdominal infection. J Surg Oncol, 2019; 119: 925–931.

12.

Shamali

, et al. Impact of obesity on short and long term results following a pancreatico-duodenectomy. Int J Surg, 2017; 42:191–196.

13.

Robinson

, et al. Peri-operative risk factors for delayed gastric emptying after a pancreaticoduodenectomy. HPB (Oxford), 2015; 17:495–501.

14.

Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: A pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 2017; 390:2627–2642.

15.

Afshin

, et al. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med, 2017; 377:13–27.

16.

Jomphe

, et al. The impact of waiting list BMI changes on the short-term outcomes of lung transplantation. Transplantation, 2018; 102:318–325.

17.

, et al. BMI as a predictor for perioperative outcome of laparoscopic colorectal surgery: A pooled analysis of comparative studies. Dis Colon Rectum, 2017; 60:433–445.

18.

Vernaci

, et al. BMI is an independent prognostic factor for late outcome in patients diagnosed with early breast cancer: A landmark survival analysis. Breast, 2019; 47:77–84.

19.

Gaujoux

, et al. Impact of obesity and body fat distribution on survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Ann Surg Oncol, 2012; 19:2908–2916.

20.

, et al. Factors influencing infectious complications after pancreatoduodenectomy. J Hepatobiliary Pancreat Sci, 2010; 17:174–179.

21.

Zarzavadjian Le Bian

, et al. Predicting the Severity of Pancreatic Fistula after Pancreaticoduodenectomy: Overweight and Blood Loss as Independent Risk Factors: Retrospective Analysis of 277 Patients. Surg Infect (Larchmt), 2019; 20:486–491.

22.

Shimizu

, et al. Influence of visceral obesity for postoperative pulmonary complications after pancreaticoduodenectomy. J Gastrointest Surg, 2011; 15:1401–1410.

23.

Dulucq

, Wintringer

, Mahajna

. Laparoscopic pancreaticoduodenectomy for benign and malignant diseases. Surg Endosc, 2006; 20:1045–1050.

24.

Bassi

, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 Years After. Surgery, 2017; 161:584–591.

25.

Wente

, et al. Postpancreatectomy hemorrhage (PPH): An international study group of pancreatic surgery (ISGPS) definition. Surgery, 2007; 142:20–25.

26.

Wente

, et al. Delayed gastric emptying (DGE) after pancreatic surgery: A suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery, 2007; 142:761–768.

27.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205–213.

28.

Mullen JT et al. Impact of body mass index on perioperative outcomes in patients undergoing major intra-abdominal cancer surgery. Ann Surg Oncol, 2008; 15:2164–2172.

29.

Hotamisligil

. Inflammation and metabolic disorders. Nature, 2006; 444:860–867.

30.

, et al. Risk factors for postoperative pancreatic fistula: Analysis of 539 successive cases of pancreaticoduodenectomy. World J Gastroenterol, 2016; 22:7797–7805.

31.

Gebauer

, et al. Options and limitations in applying the fistula classification by the International Study Group for Pancreatic Fistula. Ann Surg, 2012; 256:130–138.

32.

Andrianello

, et al. Biliary fistula after pancreaticoduodenectomy: Data from 1618 consecutive pancreaticoduodenectomies. HPB (Oxford), 2017; 19:264–269.

33.

El Nakeeb

, et al. Impact of obesity on surgical outcomes post-pancreaticoduodenectomy: A case-control study. Int J Surg, 2014; 12:488–493.

34.

Ecker

, et al. Taking theory to practice: Quality improvement for pancreaticoduodenectomy and development and integration of the fistula risk score. J Am Coll Surg, 2018; 227:430–438.e1.

35.

Schoellhammer

, Fong

, Gagandeep

. Techniques for prevention of pancreatic leak after pancreatectomy. Hepatobiliary Surg Nutr, 2014; 3:276–287.

36.

Rosso

, et al. The role of “fatty pancreas” and of BMI in the occurrence of pancreatic fistula after pancreaticoduodenectomy. J Gastrointest Surg, 2009; 13:1845–1851.

37.

Tsai

, et al. Impact of obesity on perioperative outcomes and survival following pancreaticoduodenectomy for pancreatic cancer: A large single-institution study. J Gastrointest Surg, 2010; 14:1143–1150.

38.

Kutlu

, et al. Open Pancreaticoduodenectomy case volume predicts outcome of laparoscopic approach: A population-based analysis. Ann Surg, 2018; 267:552–560.

39.

Poves

, et al. Comparison of perioperative outcomes between laparoscopic and open approach for pancreatoduodenectomy: The PADULAP randomized controlled trial. Ann Surg, 2018; 268:731–739.

40.

Michaud

, et al. Physical activity, obesity, height, and the risk of pancreatic cancer. JAMA, 2001; 286:921–929.

41.

Genkinger

, et al. A pooled analysis of 14 cohort studies of anthropometric factors and pancreatic cancer risk. Int J Cancer, 2011; 129:1708–1717.

42.

Aleman

, et al. Mechanisms of obesity-induced gastrointestinal neoplasia. Gastroenterology, 2014; 146:357–373.

43.

Ruan

, Song

, Ouyang

. Role of hypoxia in the hallmarks of human cancer. J Cell Biochem, 2009; 107:1053–1062.

44.

Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet, 2004; 363:157–163.