Author Affiliations: Istanbul Training and Research Hospital, Department of General Surgery, Istanbul 34098, Turkey (Kinaci E, Sevinc MM, Ozakay A, Bayrak S, Cakar E and Sari S)

Corresponding Author: Erdem Kinaci, MD, Org Nafiz Gurman Cad. Istanbul Egitim ve Arastirma Hastanesi, Genel Cerrahi Klinigi, Fatih, Istanbul 34098, Turkey (Tel: +90-506-508-7930; Email: erdemkinaci@gmail.com)

 

© 2016, Hepatobiliary Pancreat Dis Int. All rights reserved.

doi: 10.1016/S1499-3872(16)60068-1

Published online January 19, 2016.

 

 

Contributors: KE, SMM and OA proposed the study. KE, SMM, BS and CE collected and analyzed the data. KE, SMM, OA and SS performed the research and wrote the article. All authors contributed the study design and interpretation of the study and to further drafts. KE is the guarantor.

Funding: None.

Ethical approval: Not Needed.

Competing interest: No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

 

 

BACKGROUND: Early diagnosis of postoperative pancreatic fistula (POPF) is important for proper interventions. The preoperative, intraoperative and early postoperative biochemical markers have predictive value of POPF. The present study was to evaluate several simple biochemical parameters in the prediction of POPF.

 

METHODS: Patients who underwent pancreaticoduodenectomy in our center between 2006 and 2015 were reviewed retrospectively. Preoperative and early postoperative biochemical parameters were evaluated. Additionally, the relationship between POPF and pH and lactate level at the end of surgery were analyzed, and neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and red cell distribution width-to-platelet ratio (RPR) were calculated for postoperative days (PODs) 1 and 3. Diagnosis and grading of POPF were performed according to the standards of the International Study Group on Pancreatic Fistula. The patients were divided into two groups: Group 1 with no fistula or grade-A fistula; group 2 with grade-B or -C fistula. These simple biochemical markers were then compared between the two groups.

 

RESULTS: Serum amylase level was significantly higher at POD3, and pH level was significantly lower at the end of operation in group 2 compared with those in group 1. However, the serum amylase was below the upper limit of normal serum level and therefore, the difference was not significant in clinical practice. Receiver operating charecteristic curve analysis showed that pH level was a reliable predictor of POPF (area under the curve: 0.713; 95% CI: 0.573-0.853).

 

CONCLUSIONS: A low pH level at the end of pancreaticoduodenectomy was a risk factor of POPF. NLR, PLR, and RPR had no predictive value of POPF after pancreaticoduodenectomy.

 

(Hepatobiliary Pancreat Dis Int 2016;15:302-309)

 

KEY WORDS: acidosis; fistula; intraoperative; predictor; pancreaticoduodenectomy

Pancreaticoduodenectomy (PD) is a major abdominal operation with relatively high in-hospital mortality (up to 3%).^{[1, 2]} Postoperative pancreatic fistula (POPF) and accompanying abdominal sepsis are potential causes of mortality after pancreatic surgery.^[1-3] Prediction or early diagnosis of POPF is crucial to obviate mortality in the management of this complication. Postoperative drain fluid analysis,^[4-7] serum biochemical markers,^{[8, 9]} area of pancreatic cut surface, diameter of pancreatic canal, and texture of pancreatic remnant^{[10, 11]} are all related to POPF.

 

PD has a relatively long operation time, blood loss and potential risk of hypothermia, which are the causes of intraoperative acidosis as found in patients with major trauma.^{[1, 2, 12]} Low serum pH level is related to an increase in morbidity and mortality in trauma or major operations.^{[12, 13]}

 

Classically, neutrophilia and lymphocytopenia are expected cellular responses in systemic inflammation.^[14] Based on these data, the neutrophil-to-lymphocyte ratio (NLR) has long been used as an easy marker for inflammatory response in malignancies, chronic inflammatory diseases, and postoperative complications.^[14-16] Recently, platelet-to-lymphocyte ratio (PLR) and red cell distribution width-to-platelet ratio (RPR) have also been considered as simple inflammatory biomarkers in malignancies and inflammatory diseases for prediction of disease activity and even survival.^[17-19]

 

In this study, we aimed to evaluate intraoperative and early postoperative serum biochemical markers to predict the risk of POPF. In addition to routine biochemical parameters, intraoperative acidosis, NLR, PLR, and RPR were also evaluated.

 

 

The medical records of all patients who underwent standard or pylorus-preserving PD with or without total pancreatectomy at the authors’ institution between June 2006 and January 2015 were retrospectively reviewed using a computer-based data management system. Data reviewed included demographics, body mass index (BMI), hospital stay, intraoperative transfusions, intraoperative blood loss, operation time, the American Society of Anesthesiologists (ASA) scores, comorbities, use of prophylactic octreotid, preoperative endoscopic retrograde cholangio-pancreatography (ERCP), preoperative placement of biliary stent, preoperative endoscopic biopsies, postoperative laboratory findings, complications, and findings in physical examination. Chronic obstructive lung disease, ischemic heart disease, diabetes mellitus, hypertension, and previous major surgery were considered as comorbidities.

 

Laboratory parameters evaluated the first three days after surgery included aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), total bilirubin, amylase, albumin, calcium, hemoglobin, white blood cell (WBC), platelet, international normalized ratio (INR), and red cell distribution width (RDW). All of these parameters were measured at postoperative days 1 and 3. Corrected calcium (coCa) level was calculated according to the following formula:

 

coCa=0.8×[normal albumin (4 mg/dL)−patient’s albumin]+serum Ca

 

To indicate intraoperative acidosis, pH and lactate levels at the end of surgery were measured. Additionally, NLR, PLR, and RPR were calculated. In our case series, surgeons generally prefered to examine the C-reactive protein (CRP) level after postoperative day 3. CRP levels on postoperative day 1 or 3 were measured in only 10% of the cases. Therefore, CRP was excluded from further analysis.

 

The amylase content of drain fluids was firstly measured at postoperative day 1 and then at postoperative days 3, 5 and 7. The drainage of any measurable volume of amylase-rich fluid (3×normal serum amylase level, which was 120 U/L for our laboratory) from abdominal drains on or after postoperative day 3 was considered as an indication of POPF and was graded according to the International Study Group on Pancreatic Fistula (ISGPF).^[3] Biochemically detected leak without peripancreatic fluid collection and clinical findings of systemic inflammation was considered as a grade-A fistula. The patients with grade-A fistula do not have clinical significance. A leak with clinical findings and changes in management was considered as a grade-B fistula. A leak that required percutaneous drainage of fluid collection, reoperation, and intensive care management was considered as a grade-C fistula. The patients with grade-B or -C fistula were faced with a clinically relevant complication. According to this definition, two groups were designated: Group 1: patients with no fistula or grade-A fistula; and group 2: patients with clinically relevant POPF (grade-B or -C fistula). All of the parameters mentioned above were compared between these two groups. The measured serum levels of AST, ALT, ALP, GGT and bilirubin in the last three days preoperatively were considered as the preoperative values regardless of whether preoperative biliary drainage was applied or not. These preoperative values were also compared between the two groups because of close relationship with pathologies of the biliary system. To identify the appropriate cut-off values of significant parameters, analysis of receiver operating charecteristic (ROC) curves was performed.

 

To eliminate the effect of postoperative inflammatory processes other than POPF on the inflammatory markers, including WBC count, NLR, PLR, and RPR, the patients with an inflammatory complications other than POPF, including wound infection, pneumonia or atelectasis, were excluded. Additionally, the patients who underwent total pancreatectomy along with PD were also excluded due to the lack of pancreaticojejunostomy (PJ).

 

Pylorus-preserving PD was carried out as a routine operation and standard PD was performed in patients with the anatomic distortion of the pylorus or those with the tumor infiltration of the distal part of the stomach. End-to-side PJ or end-to-end dunking PJ was carried out by using 4/0 interrupted sutures with polypropylene or polyglactine according to the surgeons’ preferences. End-to-side hepaticojejunostomy (HJ) was performed 7 to 10 cm distally to PJ with 4/0 propylene or polydioxanone interrupted sutures. Finally, end-to-side double-layer duodenojejunostomy was performed 40 cm distally to HJ. In standard PD, gastrojejunostomy was performed 50 cm distally to HJ, and a Braun anastomosis was added. Two suction drains (Jackson-Pratt drain) were placed in each patient: one posterior to PJ and HJ and extending to the hepatorenal area, and the other on the anterior surface of the pancreatic remnant and extending to the anterior surface of the PJ. In some patients, another non-suction drain was placed in the splenorenal area according to surgeons’ preferences. In patients who had taken prophylactic octreotid according to the surgeons’ preferences, Sandostatin LAR® (Novartis AG) (0.1 mL×4; subcutaneous) was started on POD1 and continued until POD7. Low-molecular-weight heparin (Clexane® 4000 IU, Sanofi Co.) was started 1 day before the operation and continued until POD30. First-generation cephalosporins were used for perioperative antibiotic prophylaxis.

 

The SPSS 20.0 (SPSS Inc., IBM Corporation, Armonk, NY, USA) software was used for statistical analysis. Normally distributed continuous variables were expressed as mean (SD) and compared using Student’s t test. Variables not normally distributed were expressed as median (range) and compared using the Mann-Whitney U test. Nominal data were expressed as case numbers and percentages, and were compared using Fisher’s exact test.Because of the lack of data to some degree for each parameter, logistic regression analysis could not performed. The appropriate cut-off values of significant parameters were identified by using ROC curve analysis, and the sensitivity and specificity of this value were calculated. All tests were two-sided. A value of P<0.05 was considered statistically significant.

 

 

Between June 2006 and January 2015, 107 patients underwent PD at our center. Eight patients who underwent total pancreatectomy during index operation and 14 patients with early inflammatory postoperative complications other than POPF including wound infection, pneumonia, or atelectasis, were excluded. General features of the study population are summarized in Table 1. The overall POPF rate was 36.5% and the clinically relevant POPF rate (grade-B or -C) was 23.5%. The mortality rate of the entire case series decreased from 9.6% to 3.6% after 2009. Two mortal cases, one in 2008 and the other in 2013, had histories of coronary bypass surgery; one of these patients died at POD1 and the other at POD3, both were due to myocardial infarction. All of the remaining five mortalities were due to sepsis or multiple organ failure secondary to POPF.

 

Comparisons of demographics, as well as clinical and histopathological features, are shown in Table 2. Hospital stay was significantly longer in group 2. However, the other factors including age, gender, BMI, intraoperative transfusion, estimated blood loss, operation time, ASA score, comorbidities, preoperative ERCP, preoperative biliary stent placement, preoperative endoscopic biopsy and postoperative histopathological diagnosis of malignancy, were not different between the two groups. There were only three patients underwent additional resections (portal vein excision in all of the cases), therefore this parameter was not included in the analysis.

 

Univariate analysis showed that the majority of biochemical parameters were not significantly different between the two groups. Only two significantly different parameters were observed, i.e. serum amylase level at POD3 and blood pH level at the end of operation (Table 3). Since these parameters were from different time points the multivariate analysis was therefore not applicable. ROC curve analysis of the serum amylase level at POD3 for the occurrence of POPF indicated a cut-off value of 38.5 U/L [sensitivity 71%, specificity 68%; area under curve (AUC): 0.682, 95% CI: 0.537-0.827] (Fig. 1). The performance characteristics of pH level for different cut-off values according to the ROC curve analysis (AUC: 0.713, 95% CI: 0.573-0.853) (Fig. 2) are summarized in Table 4.

 

 

Although POPF is the most important complication of PD, its definition and classification is still controversial. However, the definition and grading system revealed by ISGPF^[3] have been widely accepted in recent studies. Despite the popularity of this system, different definitions have still been used in the literature; therefore, conflicting data about the rate of POPF persists. The rate of clinically relevant POPF (grade-B and -C) was 23.5% in our cases, which mirrors the upper limit of the current data in the literature. This rate varies between 14% and 23%^{[6, 8, 9, 20]} despite the presence of some very low values, such as 7% by Molinari et al.^[5]

 

Older age, male gender, high BMI, high ASA score, presence of comorbidities and surgery due to malignancy have previously been defined as patient-depending preoperative risk factors for morbidity and mortality following pancreatic surgery.^{[1, 6, 21]} However, in our cases, none of the aformentioned factors had a relationship with the increase in the risk of POPF. Additionally, fibrotic and hard texture of the pancreatic remnant, thin pancreatic canal, large volume of intraoperative blood loss and subsequent transfusions, and long operation time have been cited as intraoperative risk factors for POPF.^{[1, 5, 6, 8, 10]} The data related to the texture of the pancreatic remnant and diameter of the pancreatic canal were insufficient in our medical records; therefore, this study cannot draw any conclusion about these factors. Nevertheless, our data showed that the volume of intraoperative blood transfusion, estimated blood loss, and operation time had no relevance in the prediction of POPF.

 

Intraoperative adverse events can affect the patient outcome, especially after extensive surgical procedures. However, in the current literature, there have been conflicting results regarding the effects of intraoperative adverse events on the risk of POPF after pancreatic surgery. For instance, in two previous studies, intraoperative blood loss was shown to be a significant risk factor for clinically relevant POPF,^{[1, 11]} whereas it was not significant in the other two studies.^{[8, 9]} Similar contradictions have also been found regarding operation time and intraoperative transfusions.^{[1, 2, 22]} While long operation time was a significant risk factor for POPF according to Kimura et al^[1] and Gebauer et al,^[2] the opposite was reported by de Castro et al.^[22] In addition, other possible intraoperative adverse events related to surgical technique or anesthesia have been neglected by previous studies. Prolonged hypotension, excessive blood loss, excessive transfusions, hypothermia, long operation time, inappropriate, insufficient, or excessive fluid administration, or any adverse event appearing clinically negligible, can cause intraoperative metabolic acidosis, reflecting tissue hypoperfusion.^[23] We believe that intraoperative adverse events do not constitute a serious risk when considered separately. However, they may pose a significant risk when occuring together. Indeed, intraoperative acidosis is the “summary” of intraoperative adverse events causing tissue hypoperfusion which is an important factor effecting the healing. To our knowledge, no data have been collected about the relationship between intraoperative acidosis and the risk of fistula in gastrointestinal surgery. Our data demonstrated that a pH level lower than 7.33 at the end of surgery was significantly associated with the development of clinically relevant POPF. Intraoperative acidosis can burden the metabolic state of the patient in the early postoperative period, which is critical for healing of anastomoses. Intraoperative blood loss, long operation time, hypothermia, and inappropriate fluid administration are the aggrevaiting, but avoidable factors for metabolic acidosis. Therefore, operation team should focus on these avoidable factors to decrease the risk of POPF. Furthermore, despite the precautions, if the pH value drops below 7.33 at the end of operation in a patient underwent PD, it should be kept in mind that the risk of clinically relevant POPF is significantly higher in that patient.

 

Preoperative ERCP, biliary stent placement, or endoscopic biopsies were not related to increase in the risk of POPF in this case series. Additionally, preoperative serum level of AST, ALT, ALP, GGT, and bilirubin were also not risk factors for POPF. According to two recent studies, preoperative biliary drainage, even in patients with severe jaundice, is associated with an increase in postoperative morbidity, including wound infections, postoperative hemorrhage and metabolic complications.^{[24, 25]} In addition, de Castro et al^[22] claimed that the patients underwent PD with preoperative serum bilirubin level greater than 4.5 mg/dL (76.95 µmol/L) were asssociated with an increase in postoperative complications, in contrast to the report of Arkadopoulos et al.^[25] However, when we consider the perspective of POPF, there was no significant relationship with preoperative serum level of bilirubin and POPF in both studies. To our knowledge, only two studies have reported the relationship between preoperative AST, ALT, and ALP levels, and postoperative morbidity and mortality after PD.^{[1, 26]} The preoperative serum level of these parameters had no significant value in the prediction of morbidity and mortality, according to Winter et al.^[26] In contrast, Kimura et al^[1] reported the significant relationship between preoperative serum ALT level and mortality after PD in a very large case series including 8575 patients. However, both of these studies were not focused on the prediction of POPF in contrast to our study.

 

In our case series, NLR, PLR, and RPR measured at POD1 and POD3 have no predictive value for clinically relevant POPF. NLR has been used for more than 10 years as an inflammatory marker.^[12] It was previously reported that NLR has predictive value in gastric and pancreatic carcinoma.^{[15, 27]} In addition, Uslu et al^[19] reported that NLR and PLR were two inflammatory markers indicating disease activity in patients with rhematoid arthritis. To our knowledge, there is only one study that emphasized the predictive value of NLR on POD3 for POPF, which reported that it had no significant value.^[20] Çetinkaya et al^[18] revealed that high RPR value in patients with acute pancreatitis is associated with increased mortality. RPR was also previously used to predict the stage of hepatic fibrosis in patients with chronic hepatitits B.^[28] The present work is the first study that provides information about the predictive value of PLR and RPR for POPF, which had no significant value for this purpose. Although POPF is a severe inflammatory process, it should be noted that PD, itself, is also a source of inflammatory response. Furthermore, it is well known that neutrophilia and lymphocytopenia can develop after major surgery at the sixth hour and continuing for a period of one week.^{[14, 29]} Therefore, it can be considered that the simple inflammatory markers like NLR, PLR, and RPR may not have a significant predictive value in the early postoperative period for any inflammatory complication because of the masking effect of the inflammatory response of this major operation.

 

To predict the risk of POPF after PD, an elevated serum amylase level in the early postoperative period has been the most emphasized biochemical parameter in the literature. In our study, although there was a significant difference in serum amylase level on POD3 between the patients with clinically relevant POPF and those without POPF, the cut-off value was calculated as 38.5 U/L in the ROC curve analysis which was below the normal serum level (120 U/L for our laboratory). Therefore, it could not be defined as really “elevated”, which can complicate the clinical use of this parameter. Cloyd et al^[9] reported a strong association between serum amylase level >140 U/L on POD1 and POPF. Previously, the cut-off value had been reported as 195 U/L in a study including 51 patients.^[10] In another study, serum amylase level greater than 130 U/L on the night of surgery was reported as a significant predictor for POPF.^[8] In a relatively older study involving 1624 patients, peak serum amylase level >292 U/L within the first 48 hours after surgery was significantly associated with the risk of POPF.^[24] All of these studies revealed that the serum amylase level in the early postoperative period can have some degree of correlation with the risk of POPF. However, there has been no consensus achieved, to date, about the cut-off level.

 

The main limitation of this study is the lack of data on biochemical parameters, which made it impossible to perform more accurate statistical analyses such as logistic regression analyses. Nevertheless, we aimed to increase the reliability of our results by excluding widely insufficient parameters, including CRP, the texture of the pancreas, and the diameter of the pancreatic canal. The retrospective nature of the study can also be considered as a shortcoming, but it is very difficult to conduct a prospective controlled study about pancreatic surgery due to the very large diversity in primary pathologies, preoperative patient status, and intraoperative and postoperative factors, all of which have an effect on the risk of POPF. Thus, most of the studies addressing this issue in the literature have been retrospective in nature.

 

In conclusion, an increase in serum amylase level at POD3 after PD should alert the surgeons to the possibility of POPF. However, the cut-off value for serum amylase level in this study will not yield a practical clinical significance, as it fell within the normal range of serum level. Intraoperative acidosis can be used as a predictor for POPF. The risk of POPF is 4.46 times higher for patients with a pH level below 7.33 at the end of surgery. The simple and recently popular parameters including NLR, PLR, and RPR have no value in the prediction of POPF after PD.

 

 

1 Kimura W, Miyata H, Gotoh M, Hirai I, Kenjo A, Kitagawa Y, et al. A pancreaticoduodenectomy risk model derived from 8575 cases from a national single-race population (Japanese) using a web-based data entry system: the 30-day and in-hospital mortality rates for pancreaticoduodenectomy. Ann Surg 2014;259:773-780. PMID: 24253151

2 Gebauer F, Kloth K, Tachezy M, Vashist YK, Cataldegirmen G, Izbicki JR, et al. Options and limitations in applying the fistula classification by the International Study Group for Pancreatic Fistula. Ann Surg 2012;256:130-138. PMID: 22504279

3 Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005;138:8-13. PMID: 16003309

4 Kong J, Gananadha S, Hugh TJ, Samra JS. Pancreatoduodenectomy: role of drain fluid analysis in the management of pancreatic fistula. ANZ J Surg 2008;78:240-244. PMID: 18366393

5 Molinari E, Bassi C, Salvia R, Butturini G, Crippa S, Talamini G, 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;246:281-287. PMID: 17667507

6 Hiyoshi M, Chijiiwa K, Fujii Y, Imamura N, Nagano M, Ohuchida J. Usefulness of drain amylase, serum C-reactive protein levels and body temperature to predict postoperative pancreatic fistula after pancreaticoduodenectomy. World J Surg 2013;37:2436-2442. PMID: 23838932

7 Ansorge C, Regner S, Segersvärd R, Strömmer L. Early intraperitoneal metabolic changes and protease activation as indicators of pancreatic fistula after pancreaticoduodenectomy. Br J Surg 2012;99:104-111. PMID: 22052299

8 Palani Velu LK, Chandrabalan VV, Jabbar S, McMillan DC, McKay CJ, Carter CR, et al. Serum amylase on the night of surgery predicts clinically significant pancreatic fistula after pancreaticoduodenectomy. HPB (Oxford) 2014;16:610-619. PMID: 24246024

9 Cloyd JM, Kastenberg ZJ, Visser BC, Poultsides GA, Norton JA. Postoperative serum amylase predicts pancreatic fistula formation following pancreaticoduodenectomy. J Gastrointest Surg 2014;18:348-353. PMID: 23903930

10 Okabayashi T, Kobayashi M, Nishimori I, Sugimoto T, Onishi S, Hanazaki K. Risk factors, predictors and prevention of pancreatic fistula formation after pancreatoduodenectomy. J Hepatobiliary Pancreat Surg 2007;14:557-563. PMID: 18040620

11 Pratt WB, Callery MP, Vollmer CM Jr. Risk prediction for development of pancreatic fistula using the ISGPF classification scheme. World J Surg 2008;32:419-428. PMID: 18175170

12 Moore EE, Thomas G. Orr Memorial Lecture. Staged laparotomy for the hypothermia, acidosis, and coagulopathy syndrome. Am J Surg 1996;172:405-410. PMID: 8942535

13 Morgan K, Mansker D, Adams DB. Not just for trauma patients: damage control laparotomy in pancreatic surgery. J Gastrointest Surg 2010;14:768-772. PMID: 20224981

14 Zahorec R. Ratio of neutrophil to lymphocyte counts--rapid and simple parameter of systemic inflammation and stress in critically ill. Bratisl Lek Listy 2001;102:5-14. PMID: 11723675

15 Chen J, Hong D, Zhai Y, Shen P. Meta-analysis of associations between neutrophil-to-lymphocyte ratio and prognosis of gastric cancer. World J Surg Oncol 2015;13:122. PMID: 25889889

16 Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell 2010;140:883-899. PMID: 20303878

17 Rammohan A, Cherukuri SD, Palaniappan R, Perumal SK, Sathyanesan J, Govindan M. Preoperative platelet-lymphocyte ratio augments CA 19-9 as a predictor of malignancy in chronic calcific pancreatitis. World J Surg 2015;39:2323-2328. PMID: 25917199

18 Çetinkaya E, ?enol K, Saylam B, Tez M. Red cell distribution width to platelet ratio: new and promising prognostic marker in acute pancreatitis. World J Gastroenterol 2014;20:14450-14454. PMID: 25339831

19 Uslu AU, Küçük A, ?ahin A, Ugan Y, Y?lmaz R, Güngör T, et al. Two new inflammatory markers associated with Disease Activity Score-28 in patients with rheumatoid arthritis: neutrophil-lymphocyte ratio and platelet-lymphocyte ratio. Int J Rheum Dis 2015;18:731-735. PMID: 25900081

20 Kanda M, Fujii T, Takami H, Suenaga M, Inokawa Y, Yamada S, et al. Novel diagnostics for aggravating pancreatic fistulas at the acute phase after pancreatectomy. World J Gastroenterol 2014;20:8535-8544. PMID: 25024608

21 Parikh P, Shiloach M, Cohen ME, Bilimoria KY, Ko CY, Hall BL, et al. Pancreatectomy risk calculator: an ACS-NSQIP resource. HPB (Oxford) 2010;12:488-497. PMID: 20815858

22 de Castro SM, Busch OR, van Gulik TM, Obertop H, Gouma DJ. Incidence and management of pancreatic leakage after pancreatoduodenectomy. Br J Surg 2005;92:1117-1123. PMID: 15931656

23 Waters JH, Miller LR, Clack S, Kim JV. Cause of metabolic acidosis in prolonged surgery. Crit Care Med 1999;27:2142-2146. PMID: 10548196

24 di Mola FF, Tavano F, Rago RR, De Bonis A, Valvano MR, Andriulli A, et al. Influence of preoperative biliary drainage on surgical outcome after pancreaticoduodenectomy: single centre experience. Langenbecks Arch Surg 2014;399:649-657. PMID: 24682374

25 Arkadopoulos N, Kyriazi MA, Papanikolaou IS, Vasiliou P, Theodoraki K, Lappas C, et al. Preoperative biliary drainage of severely jaundiced patients increases morbidity of pancreaticoduodenectomy: results of a case-control study. World J Surg 2014;38:2967-2972. PMID: 24952079

26 Winter JM, Cameron JL, Yeo CJ, Alao B, Lillemoe KD, Campbell KA, et al. Biochemical markers predict morbidity and mortality after pancreaticoduodenectomy. J Am Coll Surg 2007;204:1029-1038. PMID: 17481534

27 Ben Q, An W, Wang L, Wang W, Yu L, Yuan Y. Validation of the pretreatment neutrophil-lymphocyte ratio as a predictor of overall survival in a cohort of patients with pancreatic ductal adenocarcinoma. Pancreas 2015;44:471-477. PMID: 25423560

28 Chen B, Ye B, Zhang J, Ying L, Chen Y. RDW to platelet ratio: a novel noninvasive index for predicting hepatic fibrosis and cirrhosis in chronic hepatitis B. PLoS One 2013;8:e68780. PMID: 23874760

29 Dionigi R, Dominioni L, Benevento A, Giudice G, Cuffari S, Bordone N, et al. Effects of surgical trauma of laparoscopic vs. open cholecystectomy. Hepatogastroenterology 1994;41:471-476. PMID: 7851857