Postoperative procalcitonin is a biomarker for excluding the onset of clinically relevant pancreatic fistula after pancreaticoduodenectomy

Introduction

Pancreaticoduodenectomy (PD) is still burdened by high complication rates and not-negligible mortality (up to 5%), even in high-volume centers (1). Undoubtedly, the main complication is represented by the postoperative pancreatic fistula (POPF), which often leads to other complications, as hemorrhage and sepsis that can be lethal (2).

The "International Study Group on Pancreatic Surgery (ISGPS)" defined POPF as the presence of amylase content three times higher than the Upper Limit of Normal (ULN) of serum amylase, in any measurable volume of fluid collected from abdominal drains starting from postoperative day 3 (POD3). Furthermore, based on its clinical impact, POPF is usually classified in grade A, B and C (3). Due to the lack of clinical impact, Grade A pancreatic fistula is also called “Biochemical Leak” (BL). Grade B and C POPFs are both defined “Clinically Relevant POPF” (CR-POPF) (4), due to the impact on patients’ clinical condition (5).

Even if some Authors questioned the reliability of drains amylase levels (DAL) and the utility of surgical drains, since they may not function properly, or even be the cause of POPF and infectious complications themselves (6), in other words, according to the ISGPS classification (4), DAL remain the main tool for detecting a pancreatic fistula, but its severity can properly be defined only in the aftermath on the basis of postoperative outcomes observed in weeks after the PD.

Therefore, in recent years, a radical change in management of POPF has been proposed. A transition from a reactive strategy, based on the treatment of complications when they become evident, to a proactive strategy, which aims to early detect CR-POPF and prevent further complications has been advocated (7). Hence, novel tools for early detection and prediction of CR-POPF, that recently has been reported to affect 17% (8) and up to 29% of PDs (9), are urgently needed.

Advantages of timely diagnosis and prediction of postoperative anastomotic leaks have been widely reported in different types of abdominal surgery. Recently, procalcitonin (PCT) has been suggested to be a reliable biomarker of postoperative infection (10) and anastomotic dehiscence after colorectal surgery (11), as well as a useful marker of postoperative bacterial infections associated with complications after gastrointestinal surgery (12).

The role of this biomarker in POPF assessment and prediction is still unclear. Few, and nonetheless with different and controversial conclusions, are the studies focused on this topic (13-15). To date, it has been reported that when compared to other inflammatory markers as white blood cells (WBCs) and C-reactive protein (PCR), PCT is a valid tool to predict the development of CR-POPF after PD. Unfortunately, there is no agreement on the optimal cut-off that has to be used and in which postoperative day (POD). Furthermore, in the very first postoperative days, inflammation due to the surgical trauma influences PCT values (16). Therefore, in order to improve its performance, some authors proposed the use of PCT in combination with other markers (15).

On this basis, the aim of this study was to investigate the ability of postoperative PCT, alone and in association with drains amylase levels, to predict the onset CR-POPF after pancreaticoduodenectomy. We present the following article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-22-803/rc).

Methods

From a prospective maintained database, data of consecutive patients who underwent PDs between January 2012 and January 2020 at the Department of General Surgery of the Campus Bio-Medico University of Rome, Italy, have been retrospectively analyzed.

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The local Ethical Committee approved the study (28/19 OSS ComEt CBM). Patient consent was waived due to the retrospective design of the study and considering that data are de-identified.

Demographic characteristics (age, sex, etc.), site and histological type of all tumors were collected. All patients underwent Whipple or Traverso-Longmire PD based on tumor location and received a single layer duct-to-mucosa pancreato-jejunal reconstruction. At the end of the procedure, two Jackson-Pratt abdominal drains were placed. Details of surgical procedures have been previously described (17).

Postoperative PCT and drains amylase levels concentration were measured on Alinity C and Alinity I Autoanalyzers (Abbott Park, Illinois, USA) with commercially available chemilumescence kit and immunoassays (Abbott Park, Illinois, USA). POPF was defined according to the criteria of the International Study Group of Pancreatic Surgery (3,4). The exclusion criteria are described in Figure 1.

Figure 1 Flow chart showing patients selection process. PD, pancreaticoduodenectomy.

Statistical analysis

Analysis was performed using Med-Calc 18.11.3 statistical package (MedCalc software, Mariakerke, Belgium). The “receiver operating characteristic” (ROC) curves and “Area under the curve” (AUC) defined the optimal cut-offs of PCT and DAL to identify the development of CR-POPF in different postoperative days.

Using the identified cut-offs in different postoperative days, positive predictive values (PPV) and negative predictive values (NPV) of PCT and DAL, alone and in combination with each other, were calculated to assess the probability of developing CR-POPF.

The odds ratios of the two biomarkers at their optimal cut-offs in different postoperative days were calculated to assess the risk of developing any grade of POPF (A, B and C), a CR-POPF (B and C) and grade C POPF.

All P values <0.05 were considered statistically significant.

Results

Patient’s characteristics

Demographic and clinical patients’ characteristics

Based on the criteria described in Figure 1, 130 out of 178 patients were considered eligible for the analysis. As reported in Table 1, 72 (55.4%) patients were men and 58 (44.6%) female. The median age was 71 years (range, 38–85 years), and the median BMI was 23.53 kg/m² (range, 17.2–38.5 years). Table 1 also shows comorbidities and operative data of all the series.

Five (3.8%) patients were affected by benign neoplasm and 125 (96.1%) by malignant tumors. In patients with benign neoplasm, a surgical approach was needed due to the compressive symptoms on the adjacent structures caused by the tumor.

Among patients suffering from malignant disease, 105 (80.7%) had adenocarcinoma, which turned out to be the most frequent histology in the population under study. The most frequent pathological localization was the pancreatic head, found in 90 (69.2%) patients. All others histological type and tumor sites are reported in Table 1.

CR-POPF characteristics

CR-POPF developed in 20 out of 130 patients (15.4%). More in detail, 11 (8.5%) patients developed a grade B pancreatic fistula and 9 (6.9%) a grade C pancreatic fistula.

ROC curve analysis for CR-POPF

Table 2 shows the results of ROC curves analysis and the best cut-off values for each of the biomarkers, the corresponding PPV and NPV, and the value of the AUC and the significant P values.

More in detail, we found that drains amylase levels in post-operative day 2, at the cut-offs of 2,000 U/L had the highest PPV and NPV for CR-POPF (PPV =71%, NPV =91%; P<0.001).

It should be also noted that in the same post-operative day, PCT at the cut-off of 0.5 ng/mL showed the highest NPV (91%; P<0.045) for clinically relevant pancreatic fistula and was able, when associated with DAL at the cut-off of 2,000 U/L, to increase its PPV for CR-POPF (from 71% to 81%).

In POD3, POD4 and POD5, drains amylase levels (cut-offs 780, 157 and 330 U/L, respectively) showed NPV for CR-POPF >90% (P<0.0001); furthermore, PCT at the cut-off of 0.5 ng/mL showed NPV for clinically relevant pancreatic fistula of about 90%.

Moreover, in post-operative day 5 combining DAL at the cut-off of 330 U/L and PCT at the cut-off of 0.5 ng/mL a PPV for CR-POPF of 81% was found.

Regarding the Odds ratio, in presence of DAL higher than each single cut-off in the first five postoperative days, a significant higher risk of developing CR-POPF was found. Likewise, patients who had PCT levels higher than the cut-offs from postoperative day 2 to 5 had a significant higher risk of developing CR-POPF. Notably, a progressive increased risk of developing CR-POPF from POD2 (OR =3.05; P=0.0348) to POD5 [odds ratio (OR) =4.589; P=0.0082] was observed.

An increased risk of developing CR-POPF in all five postoperative days was also found when drains amylase levels and PCT have been analyzed in combination.

As Table 2 shows, AUC values for PTC ranged from 54% in POD1 to 72% in POD5 with maximum value of 75.7% in POD4.

ROC Curve analysis for grade C POPF

In Table 3, ROC curve analysis results focused on grade C POPF have been reported showing that both drains amylase levels and PCT, alone and in combination, reached high NPV (≥92%) for grade C POPF.

Table 3 also shows significant higher risk of developing grade C CR-POPF when drains amylase levels were higher than the cut-off values on postoperative days from 1 to 5 (except in post-operative day 2). Furthermore, patients who had PCT values higher than the optimal cut-offs from POD3 to POD5 were at significant higher risk of grade C postoperative pancreatic fistula.

The association of the two biomarkers confirmed these findings. Notably, in post-operative day 2 combining DAL (cut-off of 2,000 U/L) and PCT (cut-off of 0.5 ng/mL), the risk of developing grade C pancreatic fistula was 23 times higher (OR =23.38; P=0.0037). Similarly, the association of DAL (cut-off of 300 U/L) and PCT (cut-off of 0.4 ng/mL) in POD5 showed significant higher odds of developing this fearful complication (OR =17.05; P=0.0018).

Notably, AUC values for PTC ranged from 63% in POD1 to 83% in POD5 with a maximum of 86% in POD4.

Discussion

POPF represents the main complication after PD and is often responsible for the onset of further, even lethal, complications (18). In order to improve postoperative outcomes, the availability of biomarkers able to early predict the development of CR-POPF, properly defined only in the aftermath based on postoperative outcomes observed in weeks after the PD, and to distinguish these patients from those who will not experience this complication, would be of great utility in clinical practice in order to improve outcomes.

The results of our study show that starting from POD2, drains amylase levels at different cut-offs for each postoperative day, confirmed to be a valid predictor of clinically relevant pancreatic fistula and this is consistent with what already reported in literature (19-21).

Starting from POD2, procalcitonin at the cut-off of 0.5 ng/mL showed NPV of approximately 90% for CR-POPF. Furthermore, the persistence of PCT levels ≥0.5 ng/mL from POD2 to POD5 was associated to a progressive increased risk of developing clinically relevant pancreatic fistula. ROC curve analysis confirmed the negative predicting ability of PCT for the development of CR-POPF and grade C POPF particularly in postoperative day 4 (AUC values of 75.7% and 86%, respectively).

Therefore, we can assume that when serum PCT levels are below the cut-off of 0.5 ng/mL (which, among other things, represents the maximum normal laboratory level commonly used in clinical practice) in POD2, the development of CR-POPF will most likely not occur.

Furthermore, since PCT showed highest NPV in POD2 and POD3, and considering that it has been proved the role of this biomarker in predicting abdominal infections (21,22), we could speculate that an early infectious process may lay under the development of CR-POPF.

This hypothesis could also be supported by previous experiences that reported the high accuracy of PCT in predicting severe gram-negative sepsis (23) and how the presence of gram-negative rods represents an independent risk factor for threatening Grade C POPF after PD (24). However, the lack of data about drains fluid cultures from POD2 to POD5 in our series do not allow us to confirm this hypothesis.

The high negative predictive ability of PCT was confirmed when the subgroup of grade C pancreatic fistulas was analyzed.

There are few studies in literature (13-15) that investigated relationship between post-operative PCT and development of CR-POPF. Moreover, controversial results have been reported and unique cut-off is still lacking.

In 2014, Giardino et al. conducted a prospective study on 84 patients underwent PD. In this experience, PCT in post-operative day 1 identified subjects at risk of developing CR-POPF. Patients with PCT levels higher than the identified cut-off (0.4 mg/dL) had a six times higher risk of CR-POPF (OR 5.62, 95% CI: 1.41–37.7; P=0.030) (15).

However, as already reported in literature (25), PCT increases after the first 6 hours following major abdominal surgery, regardless of any possible complications. The levels of this marker tend to decrease quickly after the first 24 hours if the inflammation disappears. One day after surgery, elevated PCT values could instead suggest the persistence of inflammation due to possible complications. Therefore, PCT assay before POD2 should not be considered reliable.

More in detail, as previously reported by other Authors (14,26), while C-reactive protein is influenced by inflammation having a half-life of 19 hours, a peak at 48 hours after the stimulation and then decrease in absence of inflammatory stimuli, PCT does not reflect the inflammatory status. PCT levels are influenced more by the presence of bacterial infectious than by other inflammatory injuries.

In disagreement with Giardino, in our experience PCT in POD1 was not associated with significant risk (OR =1.14; P=0.7844) of developing CR-POPF.

Moreover, aware that PCT is affected by surgical trauma in the first 24 postoperative hours, and considering that trauma is reduced in mini-invasive surgery, more recently Ma et al. retrospectively analyzed data of 186 patients underwent laparoscopic PD. From this analysis emerged the ability of PCT in POD3 to predict a clinically relevant pancreatic fistula (cut-off values >2.10 ng/mL, sensitivity 88.2%, specificity 92.9%, AUC =0.951; P<0.001) (14). However, it should be noted that, even in attempt to reduce the bias of inflammation due to the open surgical technique, in this study only subjects undergoing laparoscopic surgery were examined and therefore these findings cannot be generalized to all PDs.

In 2020, Zhou et al. retrospectively analyzed data of 67 patients underwent PD and 19 patients underwent distal pancreatectomy. The Authors confirmed the ability of postoperative procalcitonin to early detect CR-POPF when compared to other inflammatory markers. The best performance of procalcitonin was observed in POD1 at the cut-off of 0.67 µg/ L (sensitivity 73.68%, specificity 76.12%, AUC 0.77; 95% CI: 0.675–0.860), in POD3 at the cut-off of 0.56 µg/L (sensitivity 89.47%, specificity 64.18%, AUC 0.83; 95% CI: 0.734–0.902), and in POD5 at the cut-off of 0.46 µg/L (sensitivity 68.42%, specificity 76.12%, AUC 0.72; 95% CI: 0.621–0.818) (13). It is clear that in this experience different PCT cut-offs were identified in different postoperative days, suggesting too high variability for the clinical use of this marker. Instead, it should be emphasized that in our study PCT cut-off remained unchanged in various postoperative days, unlike what occurred with DAL.

Regarding DAL, our study confirmed its high positive predictive ability for POPF; however, as already observed by Kanda et al., it should be considered that amylase concentration in fluid collected by drains can be strongly influenced by the amount of ascitic exudate and by the effectiveness of the drains themselves. This suggests that DAL do not always increase in parallel with exacerbation of the postoperative pancreatic fistula (25,27). Being PCT levels measured on blood samples, this marker should be less influenced by non-clinical issues. Furthermore, the cut-off that emerged from our study is the maximum value of the range of normality commonly used in clinical practice.

We found interesting to investigate the association of PCT with DAL in identifying patients at higher risk of developing CR-POPF. Even Giardino et al. (15) studied this association, but they did not highlight any significant relationship with the development of clinically relevant pancreatic fistula. In our study, PCT at the cut-off of 0.5 ng/mL significantly increased the PPV for CR-POPFs of drains amylase levels in post-operative day 2 and 5. In detail, based on the results we obtained, it is likely to hypothesize that when both drains amylase levels and procalcitonin are above their respective cut-offs of 2,000 U/L and 0.5 ng/mL in POD2, patients will most likely develop a CR-POPF. A similar speculation can be made in post-operative day 5 in presence of DAL and PCT values higher than 330 U/L and 0.5 ng/mL, respectively.

Encouraged by the results obtained from this study, and with particular regard to the ability of the combination of DAL and PCT to identify subjects with extremely high (up to 23 times) risk of developing a grade C POPF, if these data would be confirmed, we could envision the use of both DAL and PCT to identify, as early as POD2, subjects who are more likely to develop severe consequences from their fistula, These high-risk patients may benefit from a more careful clinical observation and a more “intensive” postoperative management. This kind of management could include a late drains removal policy, according to the concept of their dynamic management proposed by Trudeau et al. (28), the assessment of early second level diagnostic exams (e.g., CT scan), with early percutaneous drainage of fluid collections and longer in-hospital stay, in order to decrease the failure-to-rescue rates, which according to what reported by van Dongen et al. represents a critical determinant of mortality after PD (29).

Furthermore, considering that postoperative PCT correlates with increased intra-abdominal infections after major surgery (25), this marker would represent an aid in the evaluation of those patients who could benefit from post-PD antimicrobial therapy. This therapy could be oriented on the outcome of the biliary culture intraoperatively collected (30).

Our study is not without limitations that must be identified in its retrospective nature and in the size of the sample, which even small is comparable to that reported in the few other studies published in literature and focused on this topic. Due to the above-mentioned limitations, results found in this study should be carefully considered and need larger and prospective multicenter studies to be confirmed and generalized.

Conclusions

PCT at the cut off-of 0.5 ng/mL may be used as a reliable and easily reproducible marker for identification of patients at highest risk of developing CR-POPF after PD. PCT <0.5 ng/mL could exclude the onset of this complication.

The persistence of PCT levels higher than 0.5 ng/mL from postoperative day 2 to postoperative day 5 is associated with a progressive and significant increased risk of developing this fearful complication.

Furthermore, PCT at the optimal cut-off significantly increases the predictive capacity for CR-POPF of drains amylase levels in POD2 and POD5. Therefore, the association of the two markers in these PODs could be proposed to identify patients at higher risk of CR-POPF, and in particular a grade C postoperative pancreatic fistula, that could benefit of more “intensive” postoperative management (e.g., late drains removal, targeted antimicrobial therapy, early abdominal CT-scan, etc.).

Acknowledgments

Funding: None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Journal of Gastrointestinal Oncology for the series “Pancreas Surgery”. The article has undergone external peer review.

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-22-803/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-22-803/dss

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-22-803/coif). The series “Pancreas Surgery” was commissioned by the editorial office without any funding or sponsorship. AC served as the unpaid Guest Editor of the series. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study follows the Declaration of Helsinki (as revised in 2013) and adheres to the principles of informed consent, security protection, privacy and confidentiality, and fairness and transparency. The study was approved by the Ethics Committee of University Campus Bio-Medico of Rome (28/19 OSS ComEt CBM). Patient consent was waived due to the retrospective design of the study and considering that data are de-identified.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Mintziras I, Maurer E, Kanngiesser V, et al. C-reactive protein and drain amylase accurately predict clinically relevant pancreatic fistula after partial pancreaticoduodenectomy. Int J Surg 2020;76:53-8. [Crossref] [PubMed]
2. Vollmer CM Jr, Sanchez N, Gondek S, et al. A root-cause analysis of mortality following major pancreatectomy. J Gastrointest Surg 2012;16:89-102; discussion 102-3. [Crossref] [PubMed]
3. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005;138:8-13. [Crossref] [PubMed]
4. Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 Years After. Surgery 2017;161:584-91. [Crossref] [PubMed]
5. Partelli S, Pecorelli N, Muffatti F, et al. Early Postoperative Prediction of Clinically Relevant Pancreatic Fistula after Pancreaticoduodenectomy: usefulness of C-reactive Protein. HPB (Oxford) 2017;19:580-6. [Crossref] [PubMed]
6. Huan L, Fei Q, Lin H, et al. Is peritoneal drainage essential after pancreatic surgery?: A meta-analysis and systematic review. Medicine (Baltimore) 2017;96:e9245. [Crossref] [PubMed]
7. Chen JY, Feng J, Wang XQ, et al. Risk scoring system and predictor for clinically relevant pancreatic fistula after pancreaticoduodenectomy. World J Gastroenterol 2015;21:5926-33. [Crossref] [PubMed]
8. Zhang B, Yuan Q, Li S, et al. Risk factors of clinically relevant postoperative pancreatic fistula after pancreaticoduodenectomy: A systematic review and meta-analysis. Medicine (Baltimore) 2022;101:e29757. [Crossref] [PubMed]
9. Bardol T, Delicque J, Hermida M, et al. Neck transection level and postoperative pancreatic fistula after pancreaticoduodenectomy: A retrospective cohort study of 195 patients. Int J Surg 2020;82:43-50. [Crossref] [PubMed]
10. Tan WJ, Ng WQ, Sultana R, et al. Systematic review and meta-analysis of the use of serum procalcitonin levels to predict intra-abdominal infections after colorectal surgery. Int J Colorectal Dis 2018;33:171-80. [Crossref] [PubMed]
11. Giaccaglia V, Salvi PF, Antonelli MS, et al. Procalcitonin Reveals Early Dehiscence in Colorectal Surgery: The PREDICS Study. Ann Surg 2016;263:967-72. [Crossref] [PubMed]
12. Hata T, Mizuma M, Motoi F, et al. Serum procalcitonin as an early diagnostic marker of severe postoperative complications after elective pancreaticoduodenectomy. J Hepatobiliary Pancreat Sci 2020;27:767-75. [Crossref] [PubMed]
13. Zhou Q, Xia Y, Lei Z. The predictive value of procalcitonin for postoperative early pancreatic fistula. BMC Surg 2020;20:90. [Crossref] [PubMed]
14. Ma J, Jiang P, Ji B, et al. Post-operative procalcitonin and C-reactive protein predict pancreatic fistula after laparoscopic pancreatoduodenectomy. BMC Surg 2021;21:171. [Crossref] [PubMed]
15. Giardino A, Spolverato G, Regi P, et al. C-Reactive Protein and Procalcitonin as Predictors of Postoperative Inflammatory Complications After Pancreatic Surgery. J Gastrointest Surg 2016;20:1482-92. [Crossref] [PubMed]
16. Parli SE, Trivedi G, Woodworth A, et al. Procalcitonin: Usefulness in Acute Care Surgery and Trauma. Surg Infect (Larchmt) 2018;19:131-6. [Crossref] [PubMed]
17. Caputo D, Angeletti S, Ciccozzi M, et al. Role of drain amylase levels assay and routinary postoperative day 3 abdominal CT scan in prevention of complications and management of surgical drains after pancreaticoduodenectomy. Updates Surg 2020;72:727-41. [Crossref] [PubMed]
18. Pedrazzoli S, Liessi G, Pasquali C, et al. Postoperative pancreatic fistulas: preventing severe complications and reducing reoperation and mortality rate. Ann Surg 2009;249:97-104. [Crossref] [PubMed]
19. Molinari E, Bassi C, Salvia R, et al. Amylase value in drains after pancreatic resection as predictive factor of postoperative pancreatic fistula: results of a prospective study in 137 patients. Ann Surg 2007;246:281-7. [Crossref] [PubMed]
20. Yang J, Huang Q, Wang C. Postoperative drain amylase predicts pancreatic fistula in pancreatic surgery: A systematic review and meta-analysis. Int J Surg 2015;22:38-45. [Crossref] [PubMed]
21. Israel JS, Rettammel RJ, Leverson GE, et al. Does postoperative drain amylase predict pancreatic fistula after pancreatectomy? J Am Coll Surg 2014;218:978-87. [Crossref] [PubMed]
22. Meisner M, Tschaikowsky K, Hutzler A, et al. Postoperative plasma concentrations of procalcitonin after different types of surgery. Intensive Care Med 1998;24:680-4. [Crossref] [PubMed]
23. Spoto S, Fogolari M, De Florio L, et al. Procalcitonin and MR-proAdrenomedullin combination in the etiological diagnosis and prognosis of sepsis and septic shock. Microb Pathog 2019;137:103763. [Crossref] [PubMed]
24. Chiba N, Ochiai S, Yokozuka K, et al. Risk Factors for Life-threatening Grade C Postoperative Pancreatic Fistula After Pancreatoduodenectomy Compared to Grade B. Anticancer Res 2019;39:2199-205. [Crossref] [PubMed]
25. Spoto S, Valeriani E, Caputo D, et al. The role of procalcitonin in the diagnosis of bacterial infection after major abdominal surgery: Advantage from daily measurement. Medicine (Baltimore) 2018;97:e9496. [Crossref] [PubMed]
26. Arkader R, Troster EJ, Lopes MR, et al. Procalcitonin does discriminate between sepsis and systemic inflammatory response syndrome. Arch Dis Child 2006;91:117-20. [Crossref] [PubMed]
27. Kanda M, Fujii T, Takami H, et al. Novel diagnostics for aggravating pancreatic fistulas at the acute phase after pancreatectomy. World J Gastroenterol 2014;20:8535-44. [Crossref] [PubMed]
28. Trudeau MT, Maggino L, Chen B, et al. Extended Experience with a Dynamic, Data-Driven Selective Drain Management Protocol in Pancreaticoduodenectomy: Progressive Risk Stratification for Better Practice. J Am Coll Surg 2020;230:809-818.e1. [Crossref] [PubMed]
29. van Dongen JC, Smits FJ, van Santvoort HC, et al. C-reactive protein is superior to white blood cell count for early detection of complications after pancreatoduodenectomy: a retrospective multicenter cohort study. HPB (Oxford) 2020;22:1504-12. [Crossref] [PubMed]
30. Coppola A, La Vaccara V, Farolfi T, et al. Different Biliary Microbial Flora Influence Type of Complications after Pancreaticoduodenectomy: A Single Center Retrospective Analysis. J Clin Med 2021;10:2180. [Crossref] [PubMed]