Combined inflammatory markers and amylase levels for early detection of postoperative pancreatic fistula following laparoscopic pancreaticoduodenectomy

Abstract

BACKGROUND

Clinically relevant postoperative pancreatic fistula (CR-POPF) is a major complication after laparoscopic pancreaticoduodenectomy (LPD), increasing the risks of bleeding, infection, prolonged hospitalization, and mortality. Early risk stratification remains challenging. Postoperative amylase indices and inflammatory markers may capture pancreatic leakage and systemic inflammatory response, respectively, and their combination may improve early prediction.

AIM

To evaluate the predictive value of combined inflammatory markers and amylase levels for the early detection of CR-POPF following LPD.

METHODS

This retrospective study analyzed 67 patients undergoing LPD between January 2019 and January 2023, and 31 patients developed CR-POPF. The amylase levels of drainage fluid and serum [drain fluid amylase concentration (DAC) and serum amylase concentration (SAC)] were measured on postoperative day (POD) 1 and POD3, and the drain-to-serum amylase concentration ratio (DSACR) was calculated on POD1 and POD3. Plasma interleukin-6 (IL-6), IL-8, and C-reactive protein (CRP) were also measured. Diagnostic performance was assessed using receiver operating characteristic curves, and areas under the receiver operating characteristic curves were compared using the DeLong test; multivariable logistic models were used to evaluate combined prediction.

RESULTS

Patients with CR-POPF exhibited higher SAC levels on POD1, alongside increased DAC, DSACR, and plasma IL-6, IL-8, and CRP levels on both POD1 and POD3. On POD1, DAC, SAC, and DSACR demonstrated sensitivities of 83.87%, 77.42%, and 70.97%, with specificities of 83.33%, 52.78%, and 63.89%, respectively. Plasma CRP, IL-8, and IL-6 presented sensitivities of 41.94%, 70.97%, and 64.52%, and specificities of 88.89%, 83.33%, and 88.89%, respectively. On POD3, DAC, SAC, and DSACR showed sensitivities of 93.55%, 93.55%, and 96.77%, with specificities of 61.11%, 33.33%, and 83.33%, respectively. Plasma CRP, IL-8, and IL-6 exhibited POD3 sensitivities of 70.97%, 77.42%, and 80.65%, and specificities of 80.56%, 86.11%, and 86.11%. The combination of CRP, IL-8, and IL-6 plasma levels with DAC, SAC, and DSACR demonstrated robust predictive abilities for CR-POPF prognosis on both POD1 and POD3.

CONCLUSION

This study highlights the predictive value of combined inflammatory markers and amylase levels in early CR-POPF detection after LPD, suggesting the clinical utility of these biomarkers in pancreatic surgery.

Key Words: Laparoscopic pancreaticoduodenectomy; Pancreatic fistula; Inflammatory markers; Amylase; Diagnostic accuracy

Core Tip: Clinically relevant postoperative pancreatic fistula remains a major driver of morbidity after laparoscopic pancreaticoduodenectomy. A combined model incorporating amylase indices (drain fluid amylase concentration, serum amylase concentration, drain-to-serum amylase concentration ratio) and inflammatory markers (interleukin-6/interleukin-8/C-reactive protein) showed high areas under the receiver operating characteristic curves on postoperative day 1 and postoperative day 3, supporting early identification of high-risk patients.

INTRODUCTION

Although laparoscopic pancreaticoduodenectomy (LPD), initially presented by Gagner and Pomp in 1994, has gained wider application in the treatment of periampullary and pancreatic tumors, its complex anatomy, extensive dissection, and demanding reconstruction continue to make it a highly challenging minimally invasive procedure[1,2]. Previous studies showed LPD extended operative durations and hospital stays compared to open pancreaticoduodenectomy in a comprehensive review of 10 LPD cases[3], which also showed increased postoperative complications and mortality rates[4,5]. However, its safety and feasibility has also been demonstrated in a single center and retrospective study involving 42 cases of LPD[6]. Moreover, LPD offered advantages such as reduced intraoperative blood loss, diminished transfusion needs, shorter hospital stays, and expedited postoperative recovery[8], with comparable postoperative mortality rates and long-term survival outcomes to open pancreaticoduodenectomy[7].

Severe postoperative pancreatic fistula (POPF), a postoperative complications after LPD, caused server secondary complications (e.g., postoperative bleeding, intra-abdominal abscess, delayed gastric emptying, sepsis), leading to prolonged treatment periods and surgery-related mortality[9]. According to the International Study Group into grades B and C, clinically relevant POPF (CR-POPF) poses a significant challenge in post-pancreaticoduodenectomy patient management[10]. Early identification of CR-POPF is crucial for reducing associated morbidity, mortality, and hospital readmissions[11], as well as significantly aiding clinical decision-making and improving patient outcomes[12].

Enzymatic fluid leakage from the pancreatic stump following pancreatic resection result in the increased amylase content in drain fluid, thus being a surrogate indicator for predicting POPF and postoperative acute pancreatitis[13,14]. As previous demonstrated that drain fluid amylase concentration (DAC) and serum amylase concentration (SAC) could be identified as crucial predictors for POPF[15]. An elevated systemic inflammatory response within 24 hours after surgery, reflected by inflammatory markers, such as interleukin(IL)-6, IL-8, and C-reactive protein (CRP), correlates with postoperative complications[16,17]. These inflammatory markers have emerged as potential indicators of CR-POPF following LPD[18]. Beyond absolute drain amylase values, incorporating serum amylase and ratio-based indices may improve interpretability by partially accounting for individual baseline enzyme levels and perioperative dilution effects. In parallel, systemic inflammatory activation after pancreatic surgery is driven by tissue injury, bacterial translocation, and early infectious complications[18-20]. After surgical stress, rapidly raised IL-6 precede acute-phase inflammatory responses such as CRP, which often peaks around postoperative day (POD) 2 and POD3. IL-8 reflect an amplified inflammatory milieu as its role closely linking to neutrophil chemotaxis and activation, accompanying clinically significant pancreatic leakage. Therefore, integrating amylase indices that capture local pancreatic leakage with inflammatory markers that reflect systemic response may offer complementary predictive value for early CR-POPF detection after LPD.

In this study, we evaluated whether postoperative amylase-related indices [DAC, SAC, and drain-to-serum amylase concentration ratio (DSACR)] together with systemic inflammatory markers (IL-6, IL-8, and CRP) can improve early prediction of CR-POPF after LPD. Specifically, we compared these biomarkers between patients with and without CR-POPF on POD1 and POD3, quantified their individual diagnostic performance using receiver operating characteristic (ROC) analysis, and assessed the incremental value of a combined multivariable model. We sought to establish a practical early postoperative risk-stratification approach to support timely intervention and postoperative management after pancreatic surgery.

MATERIALS AND METHODS

Subjects

The analysis encompassing subjects who underwent LPD between January 2019 and January 2023.

Inclusion criteria: (1) No evidence of distant metastasis or major vessel invasion in preoperative assessments; (2) Absence of records regarding major vessel resection or reconstruction during surgery; and (3) Availability of comprehensive preoperative, intraoperative, postoperative, and pathological data.

Exclusion criteria: (1) Incomplete or missing clinical data; (2) Simultaneous resection of other solid organs during the surgical procedure; (3) Inability to tolerate the surgery or the need for conversion to open surgery; (4) Requirement for major arterial resection or reconstruction during the procedure; (5) History of pancreatitis before surgery; and (6) Concurrent administration of other treatments like radioactive particle implantation, etc.

A total of 67 patients (45 male and 30 female) were enrolled in our study, with median age of 65 years. Among these cases, there were 21 instances of pancreatic cancer, 20 of bile duct cancer, 11 of ampullary cancer, 9 of duodenal malignancy, and 12 categorized as other types.

LPD procedure

LPD procedures involved a non-preservative pancreaticoduodenectomy was conducted laparoscopically by an experienced primary surgeon and a supporting team. Brief, patients were placed in a supine position under general anesthesia and endotracheal intubation. Established pneumoperitoneum was performed through an infraumbilical incision (2 cm). The following surgical steps included dissection to expose the inferior vena cava and mobilization of the descending duodenum walls to uncover critical blood vessels. The procedure encompassed various steps like jejunum transection, ultrasonic scalpel dissection of mesentery, stomach dissection, artery exposure, and vein closure. Additional components involved pancreas neck transection, cholecystectomy, common hepatic duct division, hepatoduodenal ligament vascularization, ultimately completing the radical pancreaticoduodenectomy. After surgery, prophylactic antibiotics were administered, secretion inhibition measures were taken, and urinary catheters were removed on the first day, promoting early mobilization.

Definition of CR-POPF

As defined by the 2016 update from the International Study Group[10], CR-POPF includes grades B and C. Grade B indicates a symptomatic fistula requiring interventions like antibiotics, drains for over 3 weeks, or percutaneous drainage. Grade C represents a symptomatic fistula associated with severe patient conditions such as sepsis, multi-organ failure necessitating intensive care unit care, and requiring aggressive surgical intervention.

Detection of inflammatory markers and amylase levels

Peripheral venous blood was POD1 and POD3. The amylase levels of drainage fluid and serum (DAC and SAC) were measured on POD1 and POD3. DSACR on POD1 was calculated by dividing the DAC on POD1 by SAC on POD1. DSACR on POD3 was calculated in the same manner. Blood samples taken in the morning and collected in ethylenediaminetetraacetic acid, heparin, and citrate-coated tubes were processed within an hour to obtain plasma via centrifugation at 2000 × g for 15 minutes at 4 degrees Celsius. IL-6 and IL-8 levels in plasma were measured using enzyme-linked immunosorbent assay kits, while plasma CRP were analyzed using a BM/Hitachi 705 instrument (Boehringer, Mannheim, Germany).

Statistical analysis

The study employed various statistical methods for data analysis: Continuous variables were presented as mean ± SD or median (interquartile range) based on their distribution, assessed using the Shapiro-Wilk test. Unpaired t-tests or Mann-Whitney U tests compared continuous variables between two groups, while Friedman analysis of variance and Dunn’s multiple comparisons test assessed changes within the same group at different time points. The analysis of categorical variables was performed by χ² or Fisher’s exact tests using GraphPad Prism 8.0 statistical software (GraphPad Software Inc., La Jolla, CA, United States). Diagnostic properties were evaluated using ROC curves generated by MedCalc software (MedCalc Software bv, Ostend, Belgium). The areas under the ROC curves (AUCs) were compared using the DeLong test[21]. Logistic regression models evaluated the association between CR-POPF (dependent variable) and predictors: (1) DAC, SAC, DSACR, and plasma levels of CRP, IL-8, IL-6; and (2) Model adequacy was assessed using the Hosmer-Lemeshow test. The area under the ROC curve of the combination of these measures was conducted. A two-sided P value of less than 0.05 was considered statistically significant.

RESULTS

Baseline characteristics of subjects

Stratifying patients based on postoperative pancreatic complications, our study encompassed a no CR-POPF group (n = 36) and a CR-POPF group (n = 31). The baseline characteristics of patients enrolled in the study were presented in the Table 1. No significant differences were observed between the no CR-POPF group and the CR-POPF group in terms of gender distribution (P = 0.142), age (P = 0.988), body mass index (P = 0.165), and pathological types (P = 0.333). Both groups exhibited similar operation times (P = 0.933) and intraoperative bleeding amounts (P = 0.420). Intraoperative transfusion showed a borderline difference between groups (yes: 12 vs 17; P = 0.089). Regarding amylase-related indices, DAC was significantly higher in the CR-POPF group on POD1 and POD3 (both P < 0.05). SAC was higher in the CR-POPF group on POD1 (P = 0.049), whereas SAC on POD3 did not differ between groups (P = 0.522). DSACR was significantly higher in the CR-POPF group on both POD1 (P = 0.022) and POD3 (P < 0.001).

Table 1 Baseline characteristics in no clinically relevant postoperative pancreatic fistula and clinically relevant postoperative pancreatic fistula groups following laparoscopic pancreaticoduodenectomy, mean ± SD/median (interquartile range).

Parameter	No CR-POPF (n = 36)	CR-POPF (n = 31)	P value
Gender			0.142
Male	19	22
Female	17	9
Age (years)	66 (58, 70.75)	64 (56, 74)	0.988
Body mass index (kg/cm2)	19.78 ± 3.25	18.65 ± 3.04	0.165
Pathological type			0.333
Pancreatic cancer	15	6
Bile duct cancer	9	11
Ampullary cancer	6	5
Duodenal malignancy	4	5
Other	2	4
Operation time (minutes)	360 (322.5, 415)	348 (300, 460)	0.933
Intraoperative bleeding (mL)	145 (100, 500)	170 (145, 280)	0.42
Intraoperative transfusion			0.089
Yes	12	17
No	24	14
Hospital stays (days)	24 (20, 30.25)	35 (27, 42)	0.001
Drain fluid amylase concentration (U/L)
POD1	2304 (1195, 3496)	7177 (4386, 11581)	< 0.001
POD3	415.5 (222.3, 3311)	1860 (1370, 2888)	0.037
Serum amylase concentration (U/L)
POD1	81.83 (23.4, 199.9)	156 (89.65, 307.8)	0.049
POD3	39.05 (25.29, 360.4)	53.21 (17.42, 86.53)	0.522
Drain-to-serum amylase concentration ratio
POD1	22.05 (10.4, 48.65)	37 (23, 48)	0.022
POD3	9.55 (5.625, 13.18)	42 (25, 80)	< 0.001

CR-POPF: Clinically relevant postoperative pancreatic fistula; POD: Postoperative day.

Comparison of plasma IL-6, IL-8, and CRP between the no CR-POPF and CR-POPF groups

On POD0, no significant differences were found in the plasma levels of IL-6, IL-8, and CRP between the no CR-POPF and CR-POPF groups (all P > 0.05; Figure 1). Conversely, on POD1 and POD3, the CR-POPF group exhibited notably higher median plasma levels of IL-6 (POD1: 264.4 pg/mL vs 181.7 pg/mL; POD3: 239.3 pg/mL vs 105.5 pg/mL), IL-8 (POD1: 110.9 pg/mL vs 62.28 pg/mL; POD3: 97.14 pg/mL vs 44.99 pg/mL), and CRP (POD1: 126.5 pg/mL vs 84.6 pg/mL; POD3: 377.5 pg/mL vs 252.7 pg/mL) compared to the no CR-POPF group (all P < 0.05). Notably, both groups showed increased plasma levels of IL-6, IL-8, and CRP on POD1 and POD3 compared to POD0 (all P < 0.05). Specifically, in the no CR-POPF group, the levels of IL-8 and CRP were higher on POD3 compared to POD1 (both P < 0.05). Meanwhile, the CRP levels in the CR-POPF group were higher on POD3 compared to POD1 (P < 0.05).

Figure 1 Comparison of plasma interleukin-6, interleukin-8, and C-reactive protein between the no clinically relevant postoperative pancreatic fistula and clinically relevant postoperative pancreatic fistula groups. CR-POPF: Clinically relevant postoperative pancreatic fistula; CRP: C-reactive protein; IL: Interleukin; POD: Postoperative day.

Diagnostic accuracy of biomarkers for CR-POPF following LPD on POD1

Table 2 and Figure 2A illustrated the performance of biomarkers on POD1. DAC exhibited a high discriminatory ability with an AUC of 0.866 (95%CI: 0.760-0.937), boasting a sensitivity of 83.87% and specificity of 83.33%. Conversely, SAC and DSACR showed AUCs of 0.641 and 0.662, sensitivitives of 77.42% and 70.97%, and specificities of 52.78% and 63.89%, respectively. Plasma CRP and IL-8 presented a moderate discriminatory capacity with an AUC of 0.658 (sensitivity: 41.94%, specificity: 88.89%) and 0.700 (sensitivity: 41.94%, specificity: 88.89%). Remarkably, plasma IL-6 showcased the highest AUC of 0.812, with a sensitivity of 64.52% and specificity of 88.89%.

Figure 2 Diagnostic accuracy of biomarkers for clinically relevant postoperative pancreatic fistula following laparoscopic pancreaticoduodenectomy. A: Postoperative day (POD) 1; B: POD3. Regarding the diagnostic accuracy for clinically relevant postoperative pancreatic fistula following laparoscopic pancreaticoduodenectomy on POD1 and POD3, several biomarkers demonstrated varying effectiveness. CRP: C-reactive protein; DAC: Drain fluid amylase concentration; DSACR: Drain-to-serum amylase concentration ratio; IL: Interleukin; POD: Postoperative day; SAC: Serum amylase concentration.

Table 2 Diagnostic accuracy of biomarkers for clinically relevant postoperative pancreatic fistula following laparoscopic pancreaticoduodenectomy on postoperative day 1.

ROC curve analysis	Drain fluid amylase concentration (U/L)	Serum amylase concentration (U/L)	Drain-to-serum amylase concentration ratio	C-reactive protein (pg/mL)	IL-8 (pg/mL)	IL-6 (pg/mL)
Areas under the ROC curve	0.866	0.641	0.662	0.658	0.708	0.812
95%CI	0.760-0.937	0.514-0.754	0.536-0.773	0.532-0.769	0.584-0.813	0.698-0.897
P value	< 0.001	0.041	0.017	0.019	0.003	< 0.001
Youden index J	0.672	0.302	0.3486	0.308	0.543	0.534
Associated criterion	> 3767	> 82.68	> 24.9	> 135.1	> 88.09	> 244.91
Sensitivity	83.87	77.42	70.97	41.94	70.97	64.52
Specificity	83.33	52.78	63.89	88.89	83.33	88.89

IL: Interleukin; ROC: Receiver operating characteristic.

Diagnostic accuracy of biomarkers for CR-POPF following LPD on POD3

As shown in Table 3 and Figure 2B, DAC and SAC exhibited AUCs of 0.648 (sensitivity: 93.55%, specificity: 61.11%) and 0.546 (sensitivity: 93.55%, specificity: 33.33%), respectively. DSACR on POD3 exhibited a higher AUC of 0.865 (95%CI: 0.760-0.936), with a sensitivity of 96.77% and specificity of 83.33%. On POD3, plasma CRP yielded an AUC of 0.790 (95%CI: 0.674-0.880) alongside a sensitivity of 70.97% and specificity of 80.56%; plasma IL-8 showed an AUC of 0.875 (95%CI: 0.772-0.943) with a sensitivity of 77.42% and specificity of 86.11%, while plasma IL-6 had the highest AUC of 0.893 (95%CI: 0.794-0.956) with 80.65% sensitivity and 86.11% specificity.

Table 3 Diagnostic accuracy of biomarkers for clinically relevant postoperative pancreatic fistula following laparoscopic pancreaticoduodenectomy on postoperative day 3.

ROC curve analysis	Serum amylase concentration (U/L)	Drain fluid amylase concentration (U/L)	Drain-to-serum amylase concentration ratio	C-reactive protein (pg/mL)	IL-8 (pg/mL)	IL-6 (pg/mL)
Areas under the ROC curve	0.648	0.546	0.865	0.790	0.875	0.893
95%CI	0.522-0.761	0.420-0.668	0.760-0.936	0.674-0.880	0.772-0.943	0.794-0.956
P value	0.043	0.525	< 0.001	< 0.001	< 0.001	< 0.001
Youden index J	0.5466	0.2688	0.8011	0.515	0.635	0.668
Associated criterion	> 734	≤ 168.36	> 14	> 336.19	> 78.23	> 177.9
Sensitivity	93.55	93.55	96.77	70.97	77.42	80.65
Specificity	61.11	33.33	83.33	80.56	86.11	86.11

IL: Interleukin; ROC: Receiver operating characteristic.

Pairwise ROC curve comparison of biomarkers for CR-POPF detection

Using the DeLong method, pairwise comparisons of AUCs for biomarkers diagnosing CR-POPF after LPD are summarized in Table 4. On POD1, DAC showed a significantly higher AUC than CRP (P = 0.017), DSACR (P = 0.007), and SAC (P = 0.001). In addition, IL-6 demonstrated a significantly higher AUC than SAC (P = 0.025). On POD3, CRP showed a significantly higher AUC than SAC (P = 0.008). Moreover, DSACR, IL-6, and IL-8 each demonstrated significantly higher AUCs than DAC (P = 0.009, 0.005, and 0.006, respectively), and DSACR, IL-6, and IL-8 also showed significantly higher AUCs than SAC (all P < 0.001). Other pairwise comparisons were not statistically significant (all P > 0.05).

Table 4 Pairwise receiver operating characteristic curve comparison of biomarkers for clinically relevant postoperative pancreatic fistula detection.

Pairwise comparison of receiver operating characteristic curves		POD1	POD3
CRP vs DAC	Difference between areas	0.208	0.142
	P value	0.017	0.118
CRP vs DSACR	Difference between areas	0.004	0.075
	P value	0.961	0.342
CRP vs IL-6	Difference between areas	0.154	0.103
	P value	0.086	0.100
CRP vs IL-8	Difference between areas	0.050	0.085
	P value	0.627	0.173
CRP vs SAC	Difference between areas	0.017	0.244
	P value	0.864	0.008
DAC vs DSACR	Difference between areas	0.203	0.217
	P value	0.007	0.009
DAC vs IL-6	Difference between areas	0.054	0.245
	P value	0.441	0.005
DAC vs IL-8	Difference between areas	0.158	0.227
	P value	0.079	0.006
DAC vs SAC	Difference between areas	0.225	0.102
	P value	0.001	0.443
DSACR vs IL-6	Difference between areas	0.150	0.028
	P value	0.123	0.671
DSACR vs IL-8	Difference between areas	0.046	0.010
	P value	0.662	0.881
DSACR vs SAC	Difference between areas	0.022	0.319
	P value	0.8625	< 0.001
IL-6 vs IL-8	Difference between areas	0.104	0.018
	P value	0.239	0.744
IL-6 vs SAC	Difference between areas	0.171	0.347
	P value	0.025	< 0.001
IL-8 vs SAC	Difference between areas	0.067	0.329
	P value	0.477	< 0.001

CRP: C-reactive protein; DAC: Drain fluid amylase concentration; DSACR: Drain-to-serum amylase concentration ratio; IL: Interleukin; POD: Postoperative day; SAC: Serum amylase concentration.

Combined diagnostic efficacy of biomarkers for CR-POPF prediction post LPD

The combined diagnostic efficacy of plasma levels of CRP, IL-8, and IL-6, in addition to DAC, SAC, and DSACR, was evaluated for predicting CR-POPF post LPD on both the POD1 and POD3 (Figure 3). According to the Hosmer-Lemeshow test, the model of biomarkers showed adequacy on both POD1 (P = 0.988) and POD3 (P = 0.725). The AUC on POD1 was 0.981 (95%CI: 0.913-0.999, P < 0.001) with a sensitivity of 96.77% and specificity of 88.89%. Similarly, on POD3, the AUC was 0.966 (95%CI: 0.890-0.995, P < 0.0001) with identical sensitivity and specificity values. These results indicate strong predictive performances for these combinations in CR-POPF prognosis.

Figure 3 The combined diagnostic efficacy of plasma levels of C-reactive protein, interleukin-8, and interleukin-6, in addition to drain fluid amylase concentration, serum amylase concentration, and drain-to-serum amylase concentration ratio, was evaluated for predicting clinically relevant postoperative pancreatic fistula post laparoscopic pancreaticoduodenectomy on both the postoperative day 1 and postoperative day 3 postoperative days. POD: Postoperative day.

DISCUSSION

Elevated levels of DAC and SAC during the early postoperative phase serve as potential indicators for the development of CR-POPF. Patients who developed CR-POPF following LPD exhibited distinctive biomarker profiles. Significant associations were observed between CR-POPF and increased SAC on POD1, elevated levels of DAC, DSACR, IL-6, IL-8, and CRP on both POD1 and POD3. Similarly, patients with CR-POPF demonstrated significantly higher plasma IL-6, IL-8, and CRP levels compared to those without CR-POPF after laparoscopic and open pancreatoduodenectomy[18]. Moreover, higher DAC and CRP levels on POD3 were specifically noted in the CR-POPF group[9].

Elevated IL-6 levels have been linked to tissue damage, initiating a cascade of inflammatory events affecting immune cell activation, endothelial dysfunction, and inflammatory cell recruitment[22]. As a pro-inflammatory marker primarily responsible for neutrophil recruitment and activation, LPD procedures increased IL-8 levels, thus leading to the severity of postoperative complications such as CR-POPF[18]. Moreover, the acute-phase protein CRP synthesized by the liver after IL-6 stimulation has widely served as a sensitive marker of systemic inflammation[23]. In our study, the elevated IL-6, IL-8, and CRP levels in patients with CR-POPF indicated a plausible association between the inflammatory response and the risk of postoperative complications.

On POD1, DAC, SAC, and DSACR demonstrated sensitivities of 83.87%, 77.42%, and 70.97%, respectively, with specificities of 83.33%, 52.78%, and 63.89%, respectively. Moreover, plasma CRP, IL-8, and IL-6 displayed sensitivities of 41.94%, 70.97%, and 64.52%, and specificities of 88.89%, 83.33%, and 88.89%, respectively. On POD3, DAC, SAC, and DSACR presented sensitivities of 93.55%, 93.55%, and 96.77%, with specificities of 61.11%, 33.33%, and 83.33%, respectively. Meanwhile, plasma CRP, IL-8, and IL-6 exhibited sensitivities of 70.97%, 77.42%, and 80.65%, and specificities of 80.56%, 86.11%, and 86.11%, respectively. Similarly, a SAC cutoff value of 70 U/L on POD1 demonstrated both sensitivity and specificity of 100% and 70%, respectively, for diagnosing CR-POPF following partial pancreaticoduodenectomy[24]. Gasteiger et al[25] found serum IL-6 on POD3 was identified as a substantial predictor for CR-POPF (AUC = 0.784, sensitivity of 63.2%, specificity of 82.6%). Moreover, DSACR exhibited sensitivity and specificity of 80.0% and 58.2% on POD1, and 77.7% and 73.3% on POD3, respectively, making it a reliable indicator for predicting CR-POPF after distal pancreatectomy[13]. Moreover, the identified cut-off values for DAC and serum CRP levels on POD3 after distal pancreatectomy for pancreatic cancer were 1206 U/L and 20.1 mg/dL, respectively, with sensitivities of 58.3% and specificities of 86.2% and 83.7%, respectively[9]. Another significant predictor, a combined score relying on POD1 values of DAC and serum CRP levels, accurately forecasted risks of CR-POPF after pancreatectomy[26]. In our study, the amalgamation of CRP, IL-8, and IL-6 plasma levels with DAC, SAC, and DSACR notably enhanced the predictive abilities for CR-POPF prognosis on both POD1 and POD3. These findings emphasized the potential of amylase levels, inflammatory markers, and their amalgamation in predicting and diagnosing CR-POPF following LPD.

However, this study has limitations, notably the relatively small sample size from a single-center study, which might restrict the generalizability of these findings. Moreover, despite the promising predictive value of these biomarkers, additional multicenter studies with larger cohorts are imperative to validate and refine their clinical utility.

CONCLUSION

The combined assessment of DAC, SAC, DSACR, IL-6, IL-8, and CRP demonstrated early identification and risk stratification for CR-POPF following LPD, being valuable tools in clinical decision-making, enabling timely interventions, and potentially improving outcomes for patients undergoing pancreatic surgery.