Nonalcoholic fatty liver disease following laparoscopic duodenum-preserving pancreatic total head resection vs laparoscopic pancreaticoduodenectomy: A retrospective cohort study

Abstract

BACKGROUND

Non-alcoholic fatty liver disease (NAFLD) is characterized by the accumulation of fat in the liver in individuals who do not consume alcohol. Several risk factors influencing the onset of NAFLD after laparoscopic pancreaticoduodenectomy (LPD) have been identified. This study investigated the risk factors associated with the development of fatty liver after laparoscopic duodenum-preserving pancreatic total head resection (LDPPHRt) and LPD.

AIM

To compare the effects of LDPPHRt and LPD on the development of postoperative NAFLD.

METHODS

This retrospective cohort study included 59 patients who were histologically diagnosed with benign or low-grade malignant pancreatic tumors and who underwent laparoscopic pancreatic surgery (LDPPHRt or LPD) between May 2020 and April 2023. Patient data on perioperative and postoperative variables were analyzed and compared. Multivariate logistic regression was used to identify pre-, peri-, and postoperative risk factors for NAFLD, with statistical significance set at P < 0.05.

RESULTS

Of the 59 patients included in the study, 17 (28.8%) developed NAFLD within 6-12 months post-surgery. The incidence of NAFLD was significantly higher in the LPD group compared to the LDPPHRt group (40.0% vs 12.5%, P = 0.022). Multivariable analysis identified the LDPPHRt surgical approach (compared to LPD) as an independent protective factor against the development of postoperative NAFLD, with an odds ratio of 0.208 (95% confidence interval: 0.046-0.931; P = 0.040).

CONCLUSION

Our findings indicate that LDPPHRt is more effective than LPD in reducing the incidence of postoperative NAFLD, which may inform surgical decision-making and optimize patient outcomes after laparoscopic pancreatic surgery.

Key Words: Non-alcoholic fatty liver disease; Laparoscopic duodenum-preserving pancreatic total head resection; Laparoscopic pancreaticoduodenectomy; Malnutrition; Insulin resistance

Core Tip: Our study demonstrates that laparoscopic pancreatic total head resection with preservation of the duodenum is more effective than laparoscopic pancreatic head duodenectomy in preventing postoperative nonalcoholic fatty liver disease. In addition, we identified the factors associated with the development of fatty liver after pancreatic surgery.

INTRODUCTION

Owing to the specificity of the pancreatic head structure, the current surgical option for most benign or low-grade malignant tumors of the pancreatic head is pancreaticoduodenectomy (PD), which is associated with a high rate of postoperative complications[1]. The concept of preserving organ function has been proposed to minimize the occurrence of postoperative complications after PD. Beger first proposed the duodenum-preserving pancreatic head resection (DPPHR) in 1972, which preserved as much as possible of the patient's normal gastrointestinal (GI) tract structure while completely resecting the lesion. The core concept of this procedure is to maximize organ preservation and ensure the structural integrity and physiological function of the digestive tract, thereby significantly reducing postoperative mortality and morbidity.

Building upon Beger's procedure, several modified techniques have been developed to enhance the safety and efficacy of the operation, such as the Frey and Berne methods[2,3]. These surgical approaches and their clinical applications in partial pancreatic head resection have laid the groundwork for the subsequent development of total pancreatic head resection. The duodenum-preserving total pancreatic head resection, proposed by Takada et al[4], preserves the duodenum and bile duct while avoiding complex digestive tract reconstruction. This refinement not only optimizes surgical techniques but also expands the surgical indications, providing patients with safer and more effective treatment options. With the continuous advancement of surgical techniques and the growing demand for minimally invasive approaches, pancreatic surgery has further evolved, the laparoscopic duodenum-preserving pancreatic total head resection (LDPPHRt) has emerged as a minimally invasive surgery[5].

As shown in Figure 1, non-alcoholic fatty liver disease (NAFLD), a distinct postoperative complication of pancreatic surgery, has been overlooked because of its perception as non-life-threatening in the clinical context[6]. Previous studies have reported the occurrence of hepatic steatosis in patients who underwent PD[7], and a case of rapidly progressive and fatal non-alcoholic steatohepatitis (NASH) following PD has been reported[8]. This finding highlights the potential severity of NAFLD in the postoperative setting and draws the attention of surgeons to the occurrence of fatty liver after pancreatic surgery. Moreover, emerging evidence suggests that metabolic abnormalities associated with fatty liver may influence the progression of pancreatic cancer. Alterations in fat metabolism and inflammatory responses within the tumor microenvironment have been specifically shown to promote pancreatic cancer progression, ultimately leading to worse patient outcomes[9]. This dual role of NAFLD, both as a postoperative complication and a potential driver of cancer progression, underscores the need for surgeons to monitor and address the occurrence of fatty liver following pancreatic surgery.

Figure 1 A 64-year-old woman received laparoscopic pancreaticoduodenectomy for chronic pancreatitis. A: Enhanced computed tomography (CT) images before pancreaticoduodenectomy (PD); B: Enhanced CT images 6 months after PD.

NAFLD, a clinicopathological syndrome, characterized by excessive intracellular fat deposition in the hepatocytes, excluding cases caused by alcohol or other well-defined liver-damaging factors. NAFLD encompasses a spectrum of conditions, including simple fatty liver, NASH, and liver cirrhosis[10]. Unlike conventional NAFLD, which is associated with obesity, de novo NAFLD after PD is thought to be associated with malnutrition due to pancreatic exocrine insufficiency or malabsorption[11]. Previous studies have reported that the incidence of and risk factors for NAFLD after PD include pancreatic cancer, positive pancreatic resection edge, high body mass index (BMI), small pancreatic volume, and long operative time[12]. However, studies on the association between NAFLD and LDPPHRt are limited.

Thus, we aimed to investigate the effect of LDPPHRt on the occurrence of postoperative fatty liver by comparing the changes in attenuation values using preoperative and postoperative liver-to-spleen enhanced computed tomography (CT) images.

MATERIALS AND METHODS

Patients

Between May 2020 and April 2023, 59 patients with a postoperative pathological diagnosis of benign or low-grade malignancy underwent laparoscopic pancreatic head resection at Wuhan Tongji Hospital. All patients were able to start drinking water 1 day after surgery and were able to start eating fluid food 3 days after surgery. Postoperative enzyme replacement therapy was initiated based on the clinical symptoms of pancreatic exocrine insufficiency, such as steatorrhea and weight loss. The initial dose was 200 mg three times daily, administered with meals. The dose was subsequently adjusted according to the severity of the steatorrhea, nutritional status, and patient response, as assessed by follow-up clinical evaluations.

Several studies have shown that new NAFLD develops within 1 year of pancreatic surgery; therefore, we employed the postoperative duration of 6-12 months as the criterion for obtaining CT images[12]. Patients with CT images obtained less than six months postoperatively and those without preoperative images were excluded. In addition, to reduce the effects of metabolic syndrome and pre-existing liver disease, patients with preoperative documented diabetes, high BMI > 28 kg/m², hepatic steatosis, and postoperative death were excluded from the study. A flow-chart is shown in Figure 2.

Figure 2 Flow chart of case screening. LPD: Laparoscopic pancreaticoduodenectomy; LDPPHRt: Laparoscopic duodenum-preserving pancreatic total head resection; NAFLD: Non-alcoholic fatty liver disease.

Surgical procedure

The LDPPHRt procedure was chosen when the bile duct and pancreaticoduodenal artery arcade could be preserved, and malignancy was excluded using routine rapid cytopathological examination of intraoperatively resected specimens. The surgical procedure has been described previously[13]. With patients under general anesthesia, five trocars were placed in a V-shape in the abdomen. After a preliminary exploration of the abdominal cavity, the gastrocolic ligament was incised to facilitate access to the gastric reticulum. The anterior fascia of the pancreas was cut, and the head of the pancreas was separated from the colonic-hepatic area to reveal the pancreas. To expose the superior mesenteric vein (SMV), the loose tissue anterior to the SMV was carefully separated, a tunnel is created anterior to the SMV and portal vein, and the pancreatic neck was dissected. The ultrasonic knife was carefully freed along the dorsal and inferior margins of the head of the pancreas to completely expose the pancreatic leptomeninges and common bile duct (CBD). An ultrasonic knife was used to separate the pancreatic tissue along the CBD from top to bottom.

The biliopancreatic jugular was entirely freed, and the main pancreatic duct was revealed at the end of the CBD, which was definitively closed or clamped shut, and then dissected. Reconstruction of the pancreatojejunostomy anastomosis was performed using Roux-en-Y jejunal collaterals.

The second employed surgical method was laparoscopic PD (LPD), using the conventional Whipple procedure. The pancreatic head, distal stomach, entire duodenum, lower end of the CBD, approximately 10 cm of the proximal jejunum, gallbladder, and regional lymph nodes were resected. Then, common hepatic duct, the pancreatic duct, and gastric stump were anastomosed to the jejunum[14].

Data collection and definition

Each patient underwent a complete physical examination, and these data were recorded. BMI, used to assess obesity, was calculated by the dividing weight (kg) by the square of the height (m²). BMI results were defined as follows: Individuals with BMI ≥ greater than or equal to 24-27.9 kg/m² were classified as overweight, and those with BMI ≥ 28 kg/m² were classified as obese[15]. The fibrosis-4 (FIB-4) index was used to predict liver fibrosis using the following formula: [Age (years) × aspartate aminotransferase (AST) (U/L)]/[platelet counts (10⁹/L)] × [alanine aminotransferase (ALT) (U/L)]^1/2[16]. NAFLD fibrosis score (NFS) was calculated using the following formula: –1.675 + 0.037 × age (years) + 0.094 × BMI (kg/m²) + 1.13 × impaired fasting glucose/diabetes (yes = 1, no = 0) + 0.99 × AST/ALT ratio - 0.013 × platelets (× 10⁹/L) – 0.66 × albumin(g/dL)[17].

Many methods can be used to diagnose hepatic steatosis, and the most accurate one is liver biopsy for histological examination; however, this invasive test has dangerous complications[18]. Some reports indicate that Portal phase-enhanced CT may offer a convenient and accessible method for quantification by measuring attenuation in Hounsfield units[19]. Therefore, in this study, we evaluated the preoperative and postoperative abdominal enhancement CT portal vein phases at least 6 months apart for the presence of postoperative NAFLD. A quantitative evaluation of the same CT slice was performed using region-of-interest (ROI) measurements. Three separate ROI measurements were performed on the liver and two separate ROI measurements on the spleen on post-contrast images to avoid artifacts and intrahepatic vascular shadows as much as possible. The CT values of each selected area were averaged, and these averaged CT values were used to determine the liver-to-spleen ratio. Fatty liver was defined as < 1 of the liver/spleen CT value ratio by the Chinese Medical Association Hepatology Branch[20]. In this study, we compared preoperative patients' backgrounds, the incidence of postoperative NAFLD, and other biochemical markers.

Ethical considerations

Ethical approval for this study was granted by the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (Approval Code: TJ-IRB202401014). Patient confidentiality and anonymity were strictly ensured, with all data anonymized before analysis. Informed consent was waived due to the retrospective nature of the study; nevertheless, stringent measures were implemented to protect patient data confidentiality.

Statistical analysis

Given the low proportion of patients lost to follow-up, a complete case analysis was performed. Furthermore, multiple imputation was used to assess the robustness of the results, and no significant changes were observed. The Shapiro-Wilk test was used to assess the normality of quantitative data. Continuous data conforming to a normal distribution were expressed as SD for the mean and analyzed using the independent sample t-test. Continuous variables that did not follow a normal distribution were reported as the median interquartile range (IQR) and compared using the Mann–Whitney U test. Categorical variables were presented as counts and percentages and analyzed using either the χ² or Fisher's exact test.

The risk factors associated with NAFLD were analyzed using univariate and multivariate (logistic regression) analyses. Only variables with a P value < 0.05 in the unifactorial analysis were included in the multifactorial analysis. All statistical analyses were performed using SPSS version 25. All statistical tests were two–sided. P values < 0.05 were considered statistically significant.

RESULTS

Based on the presence or absence of postoperative de novo NAFLD, the 59 patients were divided into two groups, and postoperative complications and biochemical data were compared. Table 1 shows a significant difference in the occurrence of postoperative fatty liver based on the choice of surgical approach in laparoscopic pancreatic head surgery (P = 0.022). To further compare the effects of the two surgical procedures on postoperative NAFLD, the baseline characteristics, including preoperative laboratory and imaging data, were compared according to the surgical method (LPD vs LDPPHRt operation).

Table 1 Postoperative laboratory and imaging variables of patients with non-alcoholic fatty liver disease or not, n (%).

Variables	Overall (n = 59)	NAFLD group (n = 17)	Not NAFLD group (n = 42)	P value
Surgical procedures				0.022
LPD	35 (59.3)	14 (40.0)	21 (60.0)
LDPPHRt	24 (40.7)	3 (12.5)	21 (87.5)
Until postoperative 1 month
Postoperative bleeding	7 (11.9)	4 (23.5)	3 (7.1)	0.176
DGE	20 (33.9)	5 (29.4)	15 (35.7)	0.643
POPF	13 (22.0)	2 (11.8)	11 (26.2)	0.388
SSI	1 (1.7)	1 (5.9)	0	0.288
Postoperative within 6-12 months
AST (U/L)	23 [17, 32]	27 [21, 37]	23 [17, 32]	0.167
ALT (U/L)	23 [13, 46]	23 [15, 49]	25 [13, 45]	0.657
AST/ALT	1.1 [0.8, 1.5]	1.3 [0.9, 1.5]	1.1 [0.7, 1.4]	0.284
FIB-4 index	1.3 [0.8, 1.9]	1.4 [0.8, 1.9]	1.1 [0.8, 1.9]	0.514
Albumin (g/L)	41.1 [37.2, 45.3]	40.5 [35.2, 43.9]	41.1 [37.5, 45.5]	0.307
Total cholesterol (mmol/L)	3.7 [3.2, 4.3]	3.9 [2.8, 4.5]	3.7 [3.3, 4.1]	0.821
Random blood glucose(mmol/L)	5.6 [5.1, 6.5]	6.0 [5.1, 9.5]	5.6 [5.1, 6.0]	0.312
TBIL (μmol/L)	10.5 [8.0, 15.0]	11.0 [8.3, 14.5]	10.2 [7.9, 15.0]	0.633
NFS index	-2.2 [-3.5, -1.3]	-2.1 [-3.7, -0.9]	-2.3 [-3.5, -1.5]	0.493
ALP (U/L)	87 [70, 129]	101 [79, 195]	86 [69, 118]	0.128

ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; DGE: Delayed gastric emptying; FIB-4: Fibrosis-4; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic pancreaticoduodenectomy; NAFLD: Nonalcoholic fatty liver disease; POPF: Postoperative pancreatic fistula; SSI: Surgical site infection; TBIL: Total bilirubin; NFS: Nonalcoholic fatty liver disease fibrosis score.

Between May 2020 and April 2023, 59 patients who underwent laparoscopic pancreatic head resection were enrolled (LPD vs LDPPHRt operation). Postoperative pathology confirmed benign or low-grade malignancies in all patients. Of the 59 patients, 39 (66.1%) were female; the median age was 53 years (IQR: 40-63); BMI was 21.3 ± 2.9 kg/m². The patients were divided into LPD and LDPPHRt groups (Table 2). In Table 2, the preoperative BMI and preoperative ALT values in the LDPPHRt group were lower than those in the LPD group, and the difference between the two groups was statistically significant (all P values < 0.05).

Table 2 Baseline characteristics of patients undergoing laparoscopic pancreaticoduodenectomy and laparoscopic duodenum-preserving total pancreatic head resection, n (%).

Variables	Overall (n = 59)	LPD group (n = 35)	LDPPHRt group (n = 24)	P value
Sex				0.628
Male	20 (33.9)	11 (31.4)	9 (37.5)
Female	39 (66.1)	24 (68.6)	15 (62.5)
Age (year)	53 (40, 63)	56 (46, 63)	52 (35, 61)	0.407
Weight (kg)	57.1 ± 9.8	59.00 ± 10.38	54.4 ± 8.4	0.078
BMI (kg/m2)	21.3 ± 2.9	21.9 ± 3.2	20.4 ± 2.2	0.042
Abdominal surgery history	23 (39.0)	12 (34.3)	11 (45.8)	0.372
ASA classification				0.181
I	2 (3.4)	1 (2.9)	1 (4.2)
II	34 (57.6)	17 (48.6)	17 (70.8)
III	23 (39.0)	17 (48.6)	6 (25.0)
Consistency of pancreas				0.334
Soft	41 (69.5)	26 (74.3)	15 (62.5)
Firm or hard	18 (30.5)	9 (25.7)	9 (37.5)
AST (U/L)	17 [14.0, 22.0]	17 [14.0, 25.0]	16 [14.0, 19.75]	0.104
ALT (U/L)	13 [10.0, 25.0]	13 [11.0, 38.0]	11.5 [9.25, 19.25]	0.042
AST/ALT	1.3 [0.9, 2.0]	1.3 [0.7, 1.5]	1.4 [1.1, 1.6]	0.083
FIB-4 index	1.2 [0.7, 1.8]	1.3 [0.8, 1.8]	1.1 [0.7, 1.8]	0.396
TC (mmol/L)	3.89 [3.44, 4.49]	3.95 [3.44, 4.50]	3.86 [3.38, 4.32]	0.611
TG (mmol/L)	0.96 [0.71, 1.40]	0.96 [0.71, 1.57]	0.97 [0.67, 1.37]	0.835
Albumin (g/L)	40.5 ± 3.3	40.0 ± 3.2	41.4 ± 3.3	0.077
Random blood glucose (mmol/L)	5.1 [4.7, 5.7]	5.4 [4.8, 6.2]	5.0 [4.5, 5.5]	0.071
Uric acid (μmol/L)	286 [250, 357]	300 [250, 364]	271 [249, 323]	0.343
Liver-to-spleen CT attenuation differential (HU)	3.8 [2.2, 7.9]	3.8 [2.3, 7.9]	3.4 [1.4, 7.9]	0.502
Liver-to-spleen CT attenuation ratio	1.03 [1.02, 1.06]	1.03 [1.02, 1.05]	1.02 [1.01, 1.06]	0.422

ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; BMI: Body mass index; CT: Computed tomography; FIB-4: Fibrosis-4; HU: Hounsfield unit; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic pancreaticoduodenectomy; TC: Total cholesterol; TG: Triglycerides.

Table 3 shows the final results in both groups. Among the enrolled patients, there were 12 with serous cystadenoma, 15 with intraductal papillary mucinous neoplasm, three with solid pseudopapillary neoplasm, seven with pancreatic neuroendocrine neoplasm, nine with chronic pancreatitis, and 13 with others. No significant differences were observed between the two groups regarding operative time, intraoperative bleeding, or days of hospitalization. Until a month postoperatively, postoperative delayed gastric emptying (DGE) was found in 20 patients, and postoperative pancreatic fistula (POPF) was found in 13 patients. Significant differences were found between the two groups in DGE (P = 0.006) and POPF (P = 0.037).

Table 3 Intra- and postoperative characteristics of patients undergoing laparoscopic pancreaticoduodenectomy and laparoscopic duodenum-preserving total pancreatic head resection, n (%).

Variables	Overall (n = 59)	LPD group (n = 35)	LDPPHRt group (n = 24)	P value
Operative time (min)	310 [270, 346]	312 [270, 360]	309 [291, 329]	0.616
Blood loss (mL)	100 [50, 200]	100 [50, 149]	125 [50, 200]	0.426
Days of hospitalization (day)	27 [23, 40]	27 [24, 43]	27 [21, 38]	0.378
Postoperative pathology				< 0.001
IPMN	15 (25.4)	6 (17.1)	9 (37.5)
SCA	12 (20.3)	2 (5.7)	10 (41.7)
SPN	3 (5.1)	2 (5.7)	1 (4.2)
PNEN	7 (11.9)	5 (14.3)	2 (8.3)
Chronic pancreatitis	9 (15.3)	9 (25.7)	0
Others	13 (22.0)	11 (31.4)	2 (8.3)
Until postoperative 1 month
Postoperative bleeding	7 (11.9)	3 (8.6)	2 (8.3)	1.000
DGE	20 (33.9)	7 (20.0)	13 (54.2)	0.006
POPF	13 (22.0)	6 (17.1)	10 (41.7)	0.037
SSI	1 (1.7)	0	1 (4.2)	0.407
ERT	28 (47.5)	15 (42.9)	13 (54.2)	0.393

DGE: Delayed gastric emptying; ERT: Enzyme replacement therapy; IPMN: Intraductal papillary mucinous neoplasm; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic pancreaticoduodenectomy; POPF: Postoperative pancreatic fistula; SSI: Surgical site infection; PNEN: Pancreatic neuroendocrine neoplasm; SCA: Serous cystadenoma; SPN: Solid pseudopapillary neoplasm.

Between 6-12 months after surgery, patients who underwent LPD showed significantly higher AST levels (P = 0.022) and FIB-4 index scores (P = 0.020) than those who underwent LDPPHRt operation. Among the 59 patients, 17 (28.8%) developed NAFLD. The incidence of NAFLD was significantly higher in the LPD group than in the LDDPHRt group (LPD, 40.0%; LDPPHRt, 12.5%). Significant differences were observed in CT attenuation (P = 0.018) and liver-to-spleen ratio (P = 0.030) (Table 4).

Table 4 Postoperative laboratory and imaging variables of patients undergoing laparoscopic pancreaticoduodenectomy and laparoscopic duodenum-preserving total pancreatic head resection, n (%).

Variables	Overall (n = 59)	LPD group (n = 35)	LDPPHRt group (n = 24)	P value
Postoperative within 6-12 months
Weight (kg)	55.17 ± 8.53	55.81 ± 8.69	54.24 ± 8.37	0.492
AST (U/L)	23 [17, 32]	27 [20, 42]	21 [16, 27]	0.022
ALT (U/L)	23 [13, 46]	28 [12, 47]	22 [13, 40]	0.346
AST/ALT	1.1 [0.8, 1.5]	1.1 [0.9, 1.5]	1.2 [0.7, 1.4]	0.459
FIB-4 index	1.3 [0.8, 1.9]	1.4 [0.8, 2.2]	0.8 [0.7, 1.6]	0.020
Albumin (g/L)	40.5 ± 3.3	40.5 ± 6.0	41.2 ± 6.0	0.645
Total cholesterol (mmol/L)	3.7 [3.2, 4.3]	3.8 [3.4, 4.4]	3.5 [3.2, 4.1]	0.203
Random blood glucose(mmol/L)	5.6 [5.1, 6.5]	5.7 [5.2, 6.6]	5.6 [4.9, 6.7]	0.534
TBIL (μmol/L)	10.5 [8.0, 15.0]	11.0 [8.7, 15.0]	9.0 [6.8, 14.5]	0.316
NFS index	-2.2 [-3.5, -1.3]	-2.0 [-3.0, -1.0]	-2.8 [-3.9, -1.8]	0.050
ALP (U/L)	87 [70.0, 129.0]	95.0 [78.0, 192.0]	84.5 [58.5, 97.0]	0.035
de novo NAFLD	17 (28.8)	14 (40.0)	3 (12.5)	0.022
Liver-to-spleen CT attenuation differential (HU)	5.0 [-3.3, 11.5]	1.7 [-8.0, 9.0]	9.4 [4.1, 14.3]	0.018
Liver-to-spleen CT attenuation ratio	1.0 [1.0, 1.1]	1.0 [0.9, 1.1]	1.1 [1.0, 1.1]	0.030

ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; CT: Computed tomography; FIB-4: Fibrosis-4; GGT: Gamma-glutamyl transferase; HU: Hounsfield unit; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic Pancreaticoduodenectomy; NAFLD: Nonalcoholic fatty liver disease; TBIL: Total bilirubin; NFS: Nonalcoholic fatty liver disease fibrosis score.

To assess the critical risk factors associated with the development of NAFLD after laparoscopic pancreatic head resection, univariate analyses were performed using various pre-, intra-, and postoperative variables (Table 5). Multivariate logistic regression analysis indicated that LDPPHRt served as an independent protective factor, whereas preoperative triglycerides (TG) emerged as an independent risk factor for de novo NAFLD within one year (Table 6).

Table 5 Univariate analysis of the risk factors for nonalcoholic fatty liver disease.

Factor	Univariate analysis
	OR (95%CI)	P value
Sex
Male	1.091 (0.334-3.563)	0.885
Female	Reference
Age
≥ 60 years	2.222 (0.694-7.118)	0.179
< 60 years	Reference
BMI
Abnormal (preoperative BMI, ≥ 25 kg/m2)	1.586 (0.334-7.530)	0.562
Normal	Reference
Consistency of pancreas
Soft	2.593 (0.641-10.488)	0.182
Firm or hard	Reference
Preoperative TC	1.538 (0.845-2.799)	0.159
Preoperative TG	5.896 (1.702-20.425)	0.005
Operation time	1.005 (0.995-1.014)	0.338
Blood loss	1.000 (0.994-1.006)	0.975
Postoperative variables
AST/ALT	1.443 (0.549-3.789)	0.457
FIB-4 index	1.271 (0.905-1.784)	0.167
Albumin	0.927 (0.840-1.024)	0.135
Total cholesterol	1.223 (0.775-1.931)	0.387
Random blood glucose	1.317 (0.948-1.830)	0.101
NFS index	1.205 (0.851-1.707)	0.293
Postoperative bleeding	0.577 (0.088-3.802)	0.568
DGE	0.917 (0.281-2.944)	0.885
POPF	3.705 (0.750-18.739)	0.107
Days of hospitalization (day)	1.023 (0.986-1.062)	0.228
Surgical procedures
LDPPHRt	0.214 (0.054-0.857)	0.029
LPD	Reference

ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; BMI: Body mass index; CI: Confidence intervals; DGE: Delayed gastric emptying; FIB-4: Fibrosis-4; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic pancreaticoduodenectomy; POPF: Postoperative pancreatic fistula; TC: Total cholesterol; TG: Triglycerides; OR: Odds ratio; NFS: Nonalcoholic fatty liver disease fibrosis score.

Table 6 Multivariate analysis of the risk factors for nonalcoholic fatty liver disease.

Factor	Multivariate analysis
	OR (95%CI)	P value
Surgical procedures
LDPPHRt	0.208 (0.046-0.931)	0.040
LPD	Reference
Preoperative TG	6.251 (1.656-23.591)	0.007

CI: Confidence intervals; LDPPHRt: Laparoscopic duodenum-preserving total pancreatic head resection; LPD: Laparoscopic pancreaticoduodenectomy; TG: Triglycerides; OR: Odds ratio.

DISCUSSION

In this study, we found that LDPPHRt significantly reduced the incidence of NAFLD compared to LPD. This reduction may be linked to the higher AST levels and FIB-4 index scores observed after LPD. Furthermore, no significant differences were observed between the two groups in terms of operative time, intraoperative bleeding, or days of hospitalization. We focused on patients who developed de novo NAFLD after surgery in both groups; our study showed an incidence of rate of 28.8%, which falls within the previously reported range of 11.9%-37.0%[21-23].

Moreover, the observed differences in DGE and POPF between the two surgical groups may have indirect implications on the development of postoperative NAFLD. In the present study, a higher incidence of DGE was observed in the LDPPHRt group. Although DGE is often considered a postoperative complication, it has been linked to the effects of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) on lipid metabolism. GLP-1 RAs slows gastric emptying and GI motility, thereby reducing glucose absorption and controlling postprandial glucose and TG fluctuations[24]. Additionally, GLP-1 RAs enhance insulin secretion from pancreatic β-cells while suppressing glucagon release from α-cells, further contributing to their glucose-lowering effects[25].

Previous studies reported that NAFLD is associated with metabolic syndromes including insulin resistance, which leads to increased hepatic fat accumulation[26]. The endocrine functions of the pancreas include the secretion of insulin and glucagon, which regulate blood glucose homeostasis. The physiological functions of insulin in the liver include inhibition of gluconeogenesis and promotion of lipid synthesis. When this occurs along with an impaired endocrine function of the pancreas, it promotes hyperinsulinemia and hyperglycemia associated with insulin resistance by increasing hepatic glucose production and decreasing glucose disposal. Excess glucose is converted into fatty acids that enter the tricarboxylic acid cycle and become substrates for hepatocellular lipogenesis[27]. In addition, the ability of insulin to inhibit adipose tissue lipolysis is diminished in insulin-resistant states, with increased production of TGs and free fatty acids and the subsequent increase in hepatic fat accumulation. Beger revealed that DPPHR has significant advantages over PD in terms of intra- and extra-pancreatic functions, and a 26-year follow-up study revealed that DPPHR has a protective effect on glucose homeostasis compared with pylorus-preserving PD and PD[28,29].

Overall, the favorable metabolic profile induced by GLP-1 RAs, including reduced caloric intake, improved glycemic control, and weight loss, supports their potential effectiveness in treating NAFLD[30]. However, the precise mechanisms underlying these benefits remain unclear, and further research is required to elucidate the relative contributions of these direct and indirect pathways.

In this study, a higher incidence of POPF was observed in the LDPPHRt group. POPF refers to the abnormal communication between the pancreatic ductal epithelium and the epithelial surface of the abdominal cavity or other hollow organs, characterized by the leakage of enzyme-rich pancreatic fluid[31]. Studies have shown that POPF can induce chronic pancreatitis, which is one of the main causes of exocrine pancreatic insufficiency (EPI)[31]. EPI is characterized by a decreased secretion of digestive enzymes, leading to postoperative diarrhea and malnutrition. Importantly, these metabolic disorders are also associated with the development of new-onset NAFLD[32].

However, despite the higher incidence of POPF in the LDPPHRt group, the risk of NAFLD was lower compared to that in the LPD group. Several factors can explain this paradox. First, the LDPPHRt group could better maintain the continuity of the digestive tract, which not only had the advantage of postoperative GI function recovery[33] but also reduced the likelihood of refractory cholangitis[34]. Beger et al[29] suggested that the loss of pancreatic function after PD is caused by the surgical removal of a greater percentage of pancreatic head tissue and the removal of the duodenum. Rubino et al[35] discovered that the presence of “incretin factor” in the proximal jejunum reduced insulin resistance, and the PD group who underwent proximal duodenal and jejunal resection, which blocked the proximal intestinal "anti-insulin" function, thereby increasing insulin resistance. Second, the DPPHR group exhibited better postoperative nutritional status, as indicated by higher serum albumin levels (4.2 vs 3.9 g/L; P = 0.003), which may mitigate the risk of NAFLD[36]. Low serum albumin levels may reflect impaired amino acid absorption and utilization, which are critical for the synthesis of mature very-low-density lipoprotein (VLDL) particles. Sakurai et al[37] found that increased VLDL-TG secretion may be an early pathophysiological manifestation of NAFLD. A possible explanation for this is the malabsorption of amino acids. Apo-B and microsomal TG are necessary for the formation of mature VLDL particles, lack of them can result in insufficient production of lipoproteins; Accordingly, TGs cannot be transported as VLDL out of the liver and are deposited therein, leading to hepatic steatosis[38]. However, in our study, the postoperative nutritional status showed no statistically significant difference between the two groups, which may be attributed to differences in patient populations or postoperative management.

In our study, there was a significant difference in the complications occurring within the first postoperative month between the two groups, with the LPD group having a lower complication rate than the LDPPHRt group. This difference may be attributed to the standardization of the surgical procedure. LPD is performed in specialized institutions using highly standardized procedures, resulting in a low rate of early postoperative surgery-related morbidities. In contrast, LDPPHRt is a developing surgical technique that has been used in only a limited number of patients and institutions. Therefore, the follow-up time for both groups should be increased to verify whether LDPPHRt is advantageous.

The FIB-4 and NFS are simple indices that combine laboratory blood test data to assess liver fibrosis. In our study, in terms of postoperative FIB-4, the LDPPHRt group had significant lower scores than the LPD group (P = 0.008), suggesting that LDPPHRt has a good performance in preventing liver fibrosis. However, a certain number of cases needs to be accumulated to verify the ability of LDPPHR to reduce the degree of liver fibrosis in the future.

Okamura et al[39] found that improved NAFLD after pancreatectomy contributes to patient prognosis. Yasukawa et al[40] revealed that digestive enzymes significantly reduced the onset of NAFLD after PD. Thus, aggressive nutritional support, including pancreatic exocrine replacement should be provided in the treatment of NAFLD.

The strength of our study was the strict exclusion criteria used, including preoperative BMI ≥ 28 kg/m² and excluding hidden preexisting NAFLD. In addition, we used the FIB-4 and NFS indices to investigate the relationship between LDPPHRt and the degree of hepatic fibrosis.

This study has several limitations. First, this was a single-center retrospective study, and the number patients with LDPPHRt included in the analysis was small. To address these limitations, future studies should aim to conduct large multicenter retrospective or prospective studies with expanded sample sizes, and a large multicenter retrospective study is needed to confirm the results. Second, this study included only LDPPHRt and LPD; future studies should incorporate additional surgical variants (e.g., partial incision-preserving pancreatic surgery) for a more comprehensive and in-depth analysis.

In this study, we used enhanced CT images of the portal venous phase, and the liver-to-spleen (L/S) CT attenuation ratio (threshold: < 1.0) was used to diagnose NAFLD. However, there is a potential diagnostic uncertainty in borderline cases. Therefore, we recommend that future studies integrate noninvasive techniques, such as FibroScan or MRI-PDFF, to improve the diagnostic accuracy and provide a more comprehensive evaluation of NAFLD[41]. Additionally, although we implemented rigorous exclusion criteria to control for other potential confounders, such as alcohol consumption, viral hepatitis, and metabolic comorbidities, detailed assessments of nutritional status and medication use were not fully incorporated. Therefore, future studies should include more comprehensive evaluations of malnutrition, drug-induced liver injury, and other confounders, to better understand their effects on NAFLD development and progression.

CONCLUSION

In conclusion, compared to LPD, LDPPHRt is more effective in preventing the development of NAFLD after surgery. This suggests that preserving the pancreatic and peripancreatic tissues during laparoscopic surgery may play a critical role in mitigating postoperative metabolic complications. However, acknowledging that this study was limited to LDPPHRt and LPD is important, and future research should incorporate additional surgical variants to provide a more comprehensive understanding of the impact of different laparoscopic pancreatic techniques on NAFLD.