- Email: [email protected]
Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification
ARTICLE IN PRESS
JID: YMSY
[m5G;June 25, 2018;8:56]
Surgery 0 0 0 (2018) 1–5
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification Salvatore Paiella1,∗, Matteo De Pastena1, Fabio Casciani, Teresa Lucia Pan, Selene Bogoni, Stefano Andrianello, Giovanni Marchegiani, Giuseppe Malleo, Claudio Bassi, Roberto Salvia General and Pancreatic Surgery Unit, Pancreas Institute, University of Verona, Policlinico GB Rossi, Verona, Italy
a r t i c l e
i n f o
Article history: Accepted 8 May 2018 Available online xxx
a b s t r a c t Background: Chyle leak is an uncommon complication after pancreatic surgery. The chyle leak incidence, definition, diagnosis, and treatment had been reported heterogeneously so far. Recently a consensus definition and grading system was published by the International Study Group for Pancreatic Surgery. This study aims to evaluate the differences in the clinical and economic burden of chyle leak applying the new definition. Methods: All data from patients who underwent pancreatic surgery for any disease from January 2014 to December 2016 were retrieved from the institutional prospective database. The 2017 International Study Group for Pancreatic Surgery definition and classification were applied. The classification was validated analyzing the differences in major complications, length of stay, and hospitalization costs. Results: A total of 945 patients was the final population. A chyle leak was reported in 43 patients (4.5%). Grade A chyle leak occurred in 10 patients (23.3%), Grade B chyle leak in 31 patients (72.1%), and Grade C chyle leak in 2 patients (4.6%). Chyle leak occurred as unique postoperative complication in 29 cases (67.4%). The economic analysis showed that the average costs of the 3 grades were 2,806, 7,150 and 15,684 euros respectively (P < .001). Furthermore, the length of stay, the rates of septic events, and major complications were significantly different among the 3 grades (P = .008, P = .004, and P < .001, respectively). Of note, we did not find any intraoperative factor associated with chyle leak. Conclusion: The present study confirms the validity of the International Study Group for Pancreatic Surgery classification of chyle leak. The 3 grades of chyle leak proposed identify reliably clinical and economical differences among the chyle leak cases. © 2018 Elsevier Inc. All rights reserved.
Introduction Chyle leak (CL) is a possible complication after pancreatic surgery. The development of CL can be caused by a direct lesion of the cisterna chyli, located at the same level of the pancreatic head and neck, and/or because of a surgical lesion of the main abdominal lymphatic vessels of this region during a pancreatic resection.1 The incidence of CL ranges from 1% to 16.3%, according to the series reported.2,3 Consistency in reporting CL is necessary for comparing studies or analyzing interventions aimed to prevent or treat CL. However, most of the series reported different methods in terms of definitions of CL and diagnostic procedures, leading to ∗ Corresponding author: University of Verona Hospital Trust, General and Pancreatic Surgery Unit, Pancreas Institute, Piazzale L.A. Scuro, 10, Verona, Italy. E-mail address: [email protected] (S. Paiella). 1 Salvatore Paiella and Matteo De Pastena share first authorship.
a high heterogeneity of results among the studies. This wide difference between centers in definition, diagnostics, and treatment of CL probably influenced the detection and the clinical and economic impact of CL on patients submitted to pancreatic surgery so far. Recently, the International Study Group on Pancreatic Surgery (ISGPS) defined a CL as a triglyceride-rich (≥110 mg/dL or ≥1.2 mmol/L) milk-like output from a drain after postoperative day 3, and 3 different grades of severity were defined.4 Grade A CL is a clinically irrelevant fistula that can be treated with a conservative approach, such as restrictions in the oral diet, and does not prolong the of hospital stay. Grade B CL implies the support of the artificial nutrition and the treatment options are drug therapy, such as octreotide, or interventional radiology. Patients experience a prolonged hospital stay due to CL. A grade C CL is present when an invasive approach is needed to treat it. In this case, only invasive
https://doi.org/10.1016/j.surg.2018.05.009 0039-6060/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
JID: YMSY 2
ARTICLE IN PRESS
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
therapies are performed (eg, lymphatic embolization or sclerosis, surgical reintervention), and mortality directly correlates with CL. The aim of the present study is to apply the recent ISGPS definition of CL, evaluating the differences in the clinical and economic burden of CL among the 3 grades proposed, using a single high-volume institutional cohort of patients submitted to pancreatic surgery. Methods Patients cohort The institutional review board approved the study. Patients who underwent pancreatic resection from 2014 to 2016 at the Unit of General and Pancreatic Surgery of the Pancreas Institute of the University of Verona, were retrieved from an electronic prospectively maintained database. When information was not available medical records were consulted. The following types of pancreatic surgery were considered: pancreaticoduodenectomy (PD), distal pancreatectomy with or without splenectomy (DP), and total pancreatectomy (TP). All patients were considered irrespective of the underlying pathology. Preoperative, intraoperative, and postoperative management The institutional surgical technique of PD has been already reported by our group.5 In case of periampullary tumors causing jaundice, patients could have preoperative biliary drains placed. The techniques of laparoscopic and robotic DP have already been described previously by our group6,7 ; the open technique was performed in the same fashion. The TP can be considered as a fusion of the 2 techniques. When possible, TP was performed en bloc; when not, 2 or more steps were needed, according to the results of the frozen section analysis. A vascular resection was performed when the portal vein or superior mesenteric vein was involved. Lymphadenectomy was carried out mostly in a standard fashion according to the ISGPS definition.1 When an intraoperative suspicion of lymph nodes metastasis was found, then other nodal stations were harvested. Thus, an extended lymphadenectomy was defined when during a PD the nodal stations #8p and/or #12a and/or #16, or when the nodal stations #7 and/or #8a and/or #9 were resected during a DP.1 While patients who underwent PD received the surgical drain following the Fistula Risk Score system,8,9 patients who underwent DP or TP always received the surgical drain. All patients received the placement of central venous catheters. The nasogastric tube was removed in the operative theater or at the time of admission in the surgical unit. The postoperative management followed the enhanced recovery after surgery protocol, with liquids on postoperative day (POD) 1, light breakfast and a semiliquid meal on POD 2, and an almost complete meal (with proteins, fats, and carbohydrates) on POD 3, if patients accepted. Abdominal drains were removed on POD 3 whenever the drainage fluid level on POD 1 was ≤5,0 0 0 U/L and the fluid appearance was not judged “worrisome” by the clinicians.10-12 Those in place after POD 3 were reassessed for amylase levels on POD 5; drains with amylase values ≤200 U/L were removed, while those with values >200 U/L and/or “worrisome” appearance were left in place and removed when appropriate. Patients are not submitted routinely to cross-sectional imaging at our institution; instead, they receive imaging only when clinical conditions require it. CL management When the output of drainage presented as a milk-like fluid, it was sampled and analyzed for the triglycerides. When the level was >110 mg/dL, from POD 3 on, a CL was diagnosed irrespective
of the levels of amylase of the fluid. The management of CL was based on several strategies, according to different clinical scenarios. If a CL was diagnosed after the start of the oral feeding, then a low-fat diet was administered at first. When this treatment was ineffective, standard total parenteral nutrition (TPN) was administered. In case of delayed oral feeding or when the Enteral Nutrition (EN) was already started then treatment was based on TPN ab initio. In cases where the conservative treatments were infective, a more aggressive treatment, such as an invasive procedure or reoperation, was performed. The output from a drain was calculated daily and each treatment continued until it became unequivocally limpid after the start of the normal oral feeding. We systematically do not treat CL with EN, and we do not reassess the triglycerides levels to confirm biochemically the resolution of a CL. Data collection Demographic and clinical details included age, gender, ASA score, body mass index (BMI), diabetes, Charlson Age-Comorbidity Index,13 jaundice, preoperative biliary drain placement, and neoadjuvant treatment. These new ISGPS definitions and classifications were adopted: postoperative pancreatic fistula (POPF), delayed gastric emptying, and postoperative pancreatic hemorrhage.14-16 The follow-up was conducted until POD 90. A non-CL-related abdominal collection was defined as any drained abdominal collection without the features of a CL or any abdominal collection discovered at imaging not suspicious for a chyle collection but treated with antibiotics due to clinical or biochemical signs of infection. Patients who received a diagnosis of CL were extracted from the whole cohort, and they were compared to the counterparts who did not develop a CL. Hospital costs were defined as the charge of part of the hospitalization, from POD 1 until discharge. We did not include the costs of surgery because of the different costs of the 3 surgical procedures considered. Thus, hospital costs included the costs of antibiotics, nutritional support, postoperative imaging study, surgical or interventional postoperative procedure, and intensive care unit admission due to CL. The cost analysis was subdivided according to the 3 categories of CL identified by the ISGPS.4 The financial settlement system is regulated by the Italian Ministry of Health. All amounts were expressed in euros. Statistical analysis Data were analyzed using SPSS version 22.0 (IBM Corp, Armonk, NY). Normally distributed data are presented as mean with standard deviation (SD), and skewed distributed data are presented as median with interquartile range. The group differences were assessed using a Chi-square test or Fisher’s exact test for categorical variables. For non-normally distributed continuous variables the Mann-Whitney U test was applied. Patients were divided into 3 categories: CL grade A, B, and C according to the ISGPS definition.4 Clinical outcomes were calculated across all grades to show distribution of the outcomes between the CL groups. Statistical significance was determined by a P-value of <.05. Results The final population was 945 patients; 558 (58.4%), 311 (32.5%), and 87 (9.1%) patients underwent PD, DP, and TP, respectively. Considering the ISGPS definition and classification of CL, a total of 43 cases (4.5%) of CL were reported. Among the cases of CL, 10 (23.3%), 31 (72.1%), and 2 (4.6%) were registered as grade A, grade B, and grade C, respectively. Table 1 shows the demographic and
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
ARTICLE IN PRESS
JID: YMSY
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
3
Table 1 Demographic data. Study population n = 956
Age (years, median, IRQ) Sex (male) BMI (kg/m2 , IRQ) Diabetes ASA score III–IV CACI > 4 Jaundice Preoperative biliary drain Neoadjuvant therapy Pancreatic resection Pancreaticoduodenectomy Distal pancreatectomy Total pancreatectomy Surgical approach Open Laparoscopic Vascular resection Extended lymphadenectomy EBL (cc, median, IRQ) Operative time (min, median, IRQ) Length of stay (days, median, IRQ)
Total n (%)
No chyle leak 913 (95.5%)
Chyle leak 43 (4.5%)
P value
63 (54–71) 493 (51.6%) 24 (22–27) 185 (19.5%) 160 (16.7%) 360 (37.7%) 353 (36.9%) 301 (31.8%) 158 (16.5%)
64 (54–71) 473 (51.8%) 24 (22–27) 179 (19.7%) 155 (17%) 349 (38.2%) 339 (37.1%) 290 (32%) 149 (16.3%)
59 (50–69) 20 (46.5%) 23 (22–27) 6 (14%) 5 (11.6%) 16 (36.5%) 14 (32.5%) 11 (25.6%) 9 (20.9%)
.032∗ .300 .441 .237 .195 .310 .342 .238 .270 .947
558 (58.4) 311 (32.5) 87 (9.1%)
532(58.3%) 298 (32.6%) 83 (9.1%)
26 (60.5%) 13 (30.2%) 4 (9.3%)
855 (89.4%) 101 (10.6%) 121 (12.7%) 354 (37.1%) 30 0 (20 0–450) 375 (275–445) 9 (7–16)
815 (89.3%) 98 (10.7%) 115 (12.6%) 333 (36.6%) 30 0 (20 0–450) 375 (276–445) 9 (7–17)
40 (93%) 3 (7%) 6 (14%) 21 (48.8%) 30 0 (20 0–50 0) 390 (321–465) 14 (11–25)
.315
.469 .073 .895 .847 <.001∗
∗ Statistically significant. BMI: body mass index; ASA: American Society of Anesthesiology; CACI: Charlson Age-Comorbidity Index; EBL: estimated blood loss; IRQ: interquartile range.
Table 2 Pathological data. Study Population n = 956
Pathology PDAC IPMN Ampullary neoplasm Cholangiocarcinoma pNET Cystic lesion Other R+ resection Tumor size Lymph node harvest Positive lymph node harvest
Total n° (%)
No chyle leak 913 (95.5%)
Chyle leak 43 (4.5%)
445 (46.5%) 104 (10.9%) 78 (8.1%) 38 (4%) 142 (14.9%) 65 (6.8%) 84 (8.8%) 221 (23.6%) 25 (19–35) 36 (26–48) 1 (0–4)
427 (46.8%) 100 (11%) 75 (8.2%) 37 (4.1%) 133 (14.5%) 59 (6.4%) 82 (9%) 213 (23.8%) 25 (20–35) 36 (26–47) 1 (0–4)
18 (41.9%) 4 (9.3%) 3 (7%) 1 (2.3%) 9 (20.9%) 6 (13.9%) 2 (4.7%) 8 (18.6%) 25 (17–45) 33 (25–55) 1 (0–5)
P value .217
.280 .592 .589 .413
∗
Statistically significant. PDAC: pancreatic ductal adenocarcinoma; pNET: pancreatic neuroendocrine tumor; IPMN: intraductal papillary mucinous neoplasm.
surgical data. There was no association with preoperative and intraoperative variables, and only a younger age was associated with CL. The CL group had a longer, more statistically significant length of stay (P < .001). Table 2 shows pathological data; there were no similarities among CLs in the underlying pathology, the amount of lymph-node harvested, or metastatic in cases of malignant diseases. Table 3 reports the demographic and surgical data of CL cases. There was a clear prevalence of grade B cases, and the grade C was the rarest (4.6% of CL cases, = 2). A preoperative biliary drain was associated with grade B cases only. CL was reported mostly after PD with a clear prevalence of grade B cases. Higher estimated blood loss and a longer operative time were associated with a grade B CL. A higher lymph-node ratio was associated with grade B CL. CL alone occurred in 29 cases (67.4%). Considering all postoperative complications, grade B CL was associated more frequently with other complications and this was statistically significant for POPF, septic events, or major complications (P < .001, P = .008, and P = .004, respectively, Table 4). No deaths were recorded in this series. Of note, we report only 2 grade C cases. As presumed, a grade C CL was associated with a longer hospitalization; the mean ICU recovery was 7 days.
Economic analysis The economic analysis showed that CL burdened the costs of hospitalization with different severity according to the newly proposed ISGPS classification of CL. The average cost of hospitalization was indeed €2,806 ± €1,420 for grade A CL, €7,150 ± €4,791 for grade B CL, and €15,684 ± €473 for grade C CL (Fig. 1, P < .001). Discussion Different from POPF, postoperative pancreatic hemorrhage, and delayed gastric emptying, CL is an uncommon complication following pancreatic surgery. Recently, analyzing retrospectively a large series of 3,324 pancreatic resections, Strobel et al reported a 10% incidence of CL. The authors also report that preexisting diabetes, surgery duration >180 minutes, distal pancreatectomy, the concomitant presence of POPF, and resection for malignancy are all independent risk factors for CL.17 This latter finding is interesting yet not surprising; currently, the neoadjuvant therapy has a large diffusion, especially in pancreatic surgery. The tissue distortion caused by chemo- and/or radiotherapy might play a role
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
ARTICLE IN PRESS
JID: YMSY 4
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
Fig. 1. Cost analysis of chyle leak.
Table 3 Demographic and surgical data of CL cases. Total chyle leakage n = 43
Age (years, mean, SD) Sex (male) BMI (kg/m2 , SD) Diabetes ASA score III-IV CACI > 4 Jaundice Preoperative biliary drain Neoadjuvant therapy Pancreatic resection Pancreaticoduodenectomy Distal pancreatectomy Total pancreatectomy Surgical approach Open Laparoscopic Vascular resection Extended lymphadenectomy EBL (cc, median, IRQ) Operation time (min, median, IRQ)
Grade A10 (23.3%)
Grade B31 (72.1%)
Grade C2 (4.6%)
P value
56.3 (13.2) 3 (30%) 24.9 (4.5) 1 (10%) 1 (10%) 360 (37.7%) 1 (10%) 0 (0%) 1 (10%)
58.1 (12.8) 15 (48.4%) 24 (4.1) 5 (16.1%) 4 (12.9%) 349 (38.2%) 12 (42.9%) 11 (35.5%) 8 (25.8%)
67.5 (4.5) 2 (100%) 27.5 (2.1) 0 (0%) 0 (0%) 16 (36.5%) 0 (0%) 0 (0%) 0 (0%)
.447 .179 .314 .749 .845 .310 .098 .049∗ .428 .016∗
3 (30%) 7 (70%) 0 (0%)
22 (71%) 5 (16.1%) 4 (12.9%)
1 (50%) 1 (50%) 0 (0%)
8 (80%) 2 (20%) 1 (10%) 2 (20%) 215 (205) 295.5 (144.6)
30 (96.8%) 1 (3.2%) 4 (12.9%) 18 (58.1%) 403.2 (226.2) 40 0.7 (10 0.2)
2 (100%) 0 (0%) 1 (50%) 1 (50%) 300 (282.8) 382.5 (60.1)
.179
.313 .085 .022∗ .032∗
∗ Statistically significant. BMI: body mass index; ASA: American Society of Anesthesiology; CACI: Charlson Age-Comorbidity Index; EBL: estimated blood loss; IRQ: interquartile range.
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
JID: YMSY
ARTICLE IN PRESS
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
5
Table 4 Postoperative complications associated with CL. Total chyle leakage n = 43
POPF Abdominal collection DGE Percutaneous drainage PPH Sepsis Clavien-Dindo score ≥ III Length of stay (days, median, IRQ) R+ resection Tumor size Lymph node harvest Positive lymph node harvest
Grade A10 (23.3%)
Grade B31 (72.1%)
Grade C2 (4.6%)
P value
0 3 0 0 0 0 0
(0%) (30%) (0%) (0%) (0%) (0%) (0%)
5 (16.1%) 18 (58.1%) 3 (9.7%) 3 (9.7%) 7 (22.6%) 9 (29%) 6 (19.4%)
1 2 1 0 0 2 2
(50%) (100%) (50%) (0%) (0%) (100%) (100%)
<.001∗ .121 .084 .536 .198 .008∗ .004∗
9.7 (2.2) 0 (0%) 50.8 (41.2) 24.5 (17.4) 0.5 (1.6)
19.8 (10) 8 (25.8%) 30.6 (20.5) 39.6 (15.9) 4.3 (7.2)
38 (12.7) 0 (0%) 35 (14.1) 27.5 (14.8) 0 (0)
<.001∗ .280 .201 .089 .030∗
∗ Statistically significant. DGE: delayed gastric emptying; POPF: postoperative pancreatic fistula; PPH: postoperative pancreatic hemorrhage.
in causing alterations in the structure of peripancreatic lymphatic vessels. Other studies suggest different predisposing factors, such as more extended lymphadenectomies,2 concomitant vascular resections,3 universal use of early enteral feeding,3,18 the manipulation of the para-aortic area and the retroperitoneal invasion,18 and the development of postoperative portal vein thrombosis.3 However, the results reported so far came from heterogeneous studies; hence no reliable conclusions could be made on the clinical impact of CL after pancreatic surgery. Finally, the ISGPS proposed a definition and a classification of CL. The present study is the first to apply this classification on a large single-center retrospective cohort of patients submitted to pancreatic surgery. In our experience, the most frequent grade was B. This is the result of our institutional treatment policy of CL, based on the early administration of TPN, rather than on the prescription of a low-fat diet. The use of TPN accelerates the resolution of the CL while the patient continues oral nutrition; this result is of paramount importance from a psychological standpoint. Our results confirm the validity of the ISGPS classification. In fact, the length of stay and the rates of major complications and septic events stratified consistently among the 3 grades. Even from an economic point of view, the new ISGPS classification depicted notable differences in the hospital costs among the 3 grades of CL, with a difference of €4,344 between grade A and grade B CL, €12,878 between grade A and grade C CL, and €8,534 between grade B and grade C CL. In our experience, the extended lymphadenectomy was not associated with a higher rate or a greater severity of CL. The analysis of the risk factors for CL was beyond the aims of this study. For the sake of completeness, we report a statistically significant association between CL and younger age only. A grade B CL was more frequently associated with preoperative biliary drain, DP, high estimated blood loss, and longer operative time. Conclusion In our experience, CL is a rare complication after pancreatic surgery, with a non-negligible impact on clinical and economic burden on the postoperative course. The newly proposed ISGPS classification identifies reliably and consistently the different clinical scenarios of CL and eliminates the previous heterogeneity of results in reporting CL after pancreatic surgery. References
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
A consensus statement by the international study group on pancreatic surgery (ISGPS). Surgery. 2014;156:591–600. Assumpcao L, Cameron JL, Wolfgang CL, Edil B, Choti MA, Herman JM, et al. Incidence and management of chyle leaks following pancreatic resection: A high volume single-center institutional experience. J Gastrointest Surg. 2008;12:1915–1923. Abu Hilal M, Layfield DM, Di Fabio F, Arregui-Fresneda I, Panagiotopoulou IG, Armstrong TH, et al. Postoperative chyle leak after major pancreatic resections in patients who receive enteral feed: Risk factors and management options. World J Surg. 2013;37:2918–2926. Besselink MG, van Rijssen LB, Bassi C, Dervenis C, Montorsi M, Adham M, et al. Definition and classification of chyle leak after pancreatic operation: A consensus statement by the International Study Group on pancreatic surgery. Surgery. 2017;161:365–372. Malleo G, Marchegiani G, Salvia R, Butturini G, Pederzoli P, Bassi C. Pancreaticoduodenectomy for pancreatic cancer: The Verona experience. Surg Today. 2011;41:463–470. Melotti G, Butturini G, Piccoli M, Casetti L, Bassi C, Mullineris B, et al. Laparoscopic distal pancreatectomy: Results on a consecutive series of 58 patients. Ann Surg. 2007;246:77–82. Butturini G, Damoli I, Crepaz L, Malleo G, Marchegiani G, Daskalaki D, et al. A prospective non-randomised single-center study comparing laparoscopic and robotic distal pancreatectomy. Surg Endosc. 2015;29:3163–3170. Callery MP, Pratt WB, Kent TS, Chaikof EL, Vollmer Jr CM. A prospectively validated clinical risk score accurately predicts pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg. 2013;216:1–14. McMillan MT, Malleo G, Bassi C, Allegrini V, Casetti L, Drebin JA, et al. Multicenter, prospective trial of selective drain management for pancreatoduodenectomy using risk stratification. Ann Surg. 2017;265:1209–1218. Bassi C, Molinari E, Malleo G, Crippa S, Butturini G, Salvia R, et al. Early versus late drain removal after standard pancreatic resections: Results of a prospective randomized trial. Ann Surg. 2010;252:207–214. Salvia R, Malleo G, Butturini G, Dal Molin M, Esposito A, Marchegiani G, et al. Perioperative management of patients undergoing pancreatic resection: Implementation of a care plan in a tertiary-care center. J Surg Oncol. 2013;107:51–57. 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. Chang CM, Yin WY, Wei CK, Wu CC, Su YC, Yu CH, et al. Correction: Adjusted age-adjusted Charlson comorbidity index score as a risk measure of perioperative mortality before cancer surgery. PLoS One. 2016;11. Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;161:584–591. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al. Postpancreatectomy hemorrhage (PPH): An International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery. 2007;142:20–25. Wente MN, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, et al. Delayed gastric emptying (DGE) after pancreatic surgery: A suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142:761–768. Strobel O, Brangs S, Hinz U, Pausch T, Huttner FJ, Diener MK, et al. Incidence, risk factors and clinical implications of chyle leak after pancreatic surgery. Br J Surg. 2017;104:108–117. Kuboki S, Shimizu H, Yoshidome H, Ohtsuka M, Kato A, Yoshitomi H, et al. Chylous ascites after hepatopancreatobiliary surgery. Br J Surg. 2013;100:522–527.
1. Tol JA, Gouma DJ, Bassi C, Dervenis C, Montorsi M, Adham M, et al. Definition of a standard lymphadenectomy in surgery for pancreatic ductal adenocarcinoma:
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
JID: YMSY
[m5G;June 25, 2018;8:56]
Surgery 0 0 0 (2018) 1–5
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification Salvatore Paiella1,∗, Matteo De Pastena1, Fabio Casciani, Teresa Lucia Pan, Selene Bogoni, Stefano Andrianello, Giovanni Marchegiani, Giuseppe Malleo, Claudio Bassi, Roberto Salvia General and Pancreatic Surgery Unit, Pancreas Institute, University of Verona, Policlinico GB Rossi, Verona, Italy
a r t i c l e
i n f o
Article history: Accepted 8 May 2018 Available online xxx
a b s t r a c t Background: Chyle leak is an uncommon complication after pancreatic surgery. The chyle leak incidence, definition, diagnosis, and treatment had been reported heterogeneously so far. Recently a consensus definition and grading system was published by the International Study Group for Pancreatic Surgery. This study aims to evaluate the differences in the clinical and economic burden of chyle leak applying the new definition. Methods: All data from patients who underwent pancreatic surgery for any disease from January 2014 to December 2016 were retrieved from the institutional prospective database. The 2017 International Study Group for Pancreatic Surgery definition and classification were applied. The classification was validated analyzing the differences in major complications, length of stay, and hospitalization costs. Results: A total of 945 patients was the final population. A chyle leak was reported in 43 patients (4.5%). Grade A chyle leak occurred in 10 patients (23.3%), Grade B chyle leak in 31 patients (72.1%), and Grade C chyle leak in 2 patients (4.6%). Chyle leak occurred as unique postoperative complication in 29 cases (67.4%). The economic analysis showed that the average costs of the 3 grades were 2,806, 7,150 and 15,684 euros respectively (P < .001). Furthermore, the length of stay, the rates of septic events, and major complications were significantly different among the 3 grades (P = .008, P = .004, and P < .001, respectively). Of note, we did not find any intraoperative factor associated with chyle leak. Conclusion: The present study confirms the validity of the International Study Group for Pancreatic Surgery classification of chyle leak. The 3 grades of chyle leak proposed identify reliably clinical and economical differences among the chyle leak cases. © 2018 Elsevier Inc. All rights reserved.
Introduction Chyle leak (CL) is a possible complication after pancreatic surgery. The development of CL can be caused by a direct lesion of the cisterna chyli, located at the same level of the pancreatic head and neck, and/or because of a surgical lesion of the main abdominal lymphatic vessels of this region during a pancreatic resection.1 The incidence of CL ranges from 1% to 16.3%, according to the series reported.2,3 Consistency in reporting CL is necessary for comparing studies or analyzing interventions aimed to prevent or treat CL. However, most of the series reported different methods in terms of definitions of CL and diagnostic procedures, leading to ∗ Corresponding author: University of Verona Hospital Trust, General and Pancreatic Surgery Unit, Pancreas Institute, Piazzale L.A. Scuro, 10, Verona, Italy. E-mail address: [email protected] (S. Paiella). 1 Salvatore Paiella and Matteo De Pastena share first authorship.
a high heterogeneity of results among the studies. This wide difference between centers in definition, diagnostics, and treatment of CL probably influenced the detection and the clinical and economic impact of CL on patients submitted to pancreatic surgery so far. Recently, the International Study Group on Pancreatic Surgery (ISGPS) defined a CL as a triglyceride-rich (≥110 mg/dL or ≥1.2 mmol/L) milk-like output from a drain after postoperative day 3, and 3 different grades of severity were defined.4 Grade A CL is a clinically irrelevant fistula that can be treated with a conservative approach, such as restrictions in the oral diet, and does not prolong the of hospital stay. Grade B CL implies the support of the artificial nutrition and the treatment options are drug therapy, such as octreotide, or interventional radiology. Patients experience a prolonged hospital stay due to CL. A grade C CL is present when an invasive approach is needed to treat it. In this case, only invasive
https://doi.org/10.1016/j.surg.2018.05.009 0039-6060/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
JID: YMSY 2
ARTICLE IN PRESS
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
therapies are performed (eg, lymphatic embolization or sclerosis, surgical reintervention), and mortality directly correlates with CL. The aim of the present study is to apply the recent ISGPS definition of CL, evaluating the differences in the clinical and economic burden of CL among the 3 grades proposed, using a single high-volume institutional cohort of patients submitted to pancreatic surgery. Methods Patients cohort The institutional review board approved the study. Patients who underwent pancreatic resection from 2014 to 2016 at the Unit of General and Pancreatic Surgery of the Pancreas Institute of the University of Verona, were retrieved from an electronic prospectively maintained database. When information was not available medical records were consulted. The following types of pancreatic surgery were considered: pancreaticoduodenectomy (PD), distal pancreatectomy with or without splenectomy (DP), and total pancreatectomy (TP). All patients were considered irrespective of the underlying pathology. Preoperative, intraoperative, and postoperative management The institutional surgical technique of PD has been already reported by our group.5 In case of periampullary tumors causing jaundice, patients could have preoperative biliary drains placed. The techniques of laparoscopic and robotic DP have already been described previously by our group6,7 ; the open technique was performed in the same fashion. The TP can be considered as a fusion of the 2 techniques. When possible, TP was performed en bloc; when not, 2 or more steps were needed, according to the results of the frozen section analysis. A vascular resection was performed when the portal vein or superior mesenteric vein was involved. Lymphadenectomy was carried out mostly in a standard fashion according to the ISGPS definition.1 When an intraoperative suspicion of lymph nodes metastasis was found, then other nodal stations were harvested. Thus, an extended lymphadenectomy was defined when during a PD the nodal stations #8p and/or #12a and/or #16, or when the nodal stations #7 and/or #8a and/or #9 were resected during a DP.1 While patients who underwent PD received the surgical drain following the Fistula Risk Score system,8,9 patients who underwent DP or TP always received the surgical drain. All patients received the placement of central venous catheters. The nasogastric tube was removed in the operative theater or at the time of admission in the surgical unit. The postoperative management followed the enhanced recovery after surgery protocol, with liquids on postoperative day (POD) 1, light breakfast and a semiliquid meal on POD 2, and an almost complete meal (with proteins, fats, and carbohydrates) on POD 3, if patients accepted. Abdominal drains were removed on POD 3 whenever the drainage fluid level on POD 1 was ≤5,0 0 0 U/L and the fluid appearance was not judged “worrisome” by the clinicians.10-12 Those in place after POD 3 were reassessed for amylase levels on POD 5; drains with amylase values ≤200 U/L were removed, while those with values >200 U/L and/or “worrisome” appearance were left in place and removed when appropriate. Patients are not submitted routinely to cross-sectional imaging at our institution; instead, they receive imaging only when clinical conditions require it. CL management When the output of drainage presented as a milk-like fluid, it was sampled and analyzed for the triglycerides. When the level was >110 mg/dL, from POD 3 on, a CL was diagnosed irrespective
of the levels of amylase of the fluid. The management of CL was based on several strategies, according to different clinical scenarios. If a CL was diagnosed after the start of the oral feeding, then a low-fat diet was administered at first. When this treatment was ineffective, standard total parenteral nutrition (TPN) was administered. In case of delayed oral feeding or when the Enteral Nutrition (EN) was already started then treatment was based on TPN ab initio. In cases where the conservative treatments were infective, a more aggressive treatment, such as an invasive procedure or reoperation, was performed. The output from a drain was calculated daily and each treatment continued until it became unequivocally limpid after the start of the normal oral feeding. We systematically do not treat CL with EN, and we do not reassess the triglycerides levels to confirm biochemically the resolution of a CL. Data collection Demographic and clinical details included age, gender, ASA score, body mass index (BMI), diabetes, Charlson Age-Comorbidity Index,13 jaundice, preoperative biliary drain placement, and neoadjuvant treatment. These new ISGPS definitions and classifications were adopted: postoperative pancreatic fistula (POPF), delayed gastric emptying, and postoperative pancreatic hemorrhage.14-16 The follow-up was conducted until POD 90. A non-CL-related abdominal collection was defined as any drained abdominal collection without the features of a CL or any abdominal collection discovered at imaging not suspicious for a chyle collection but treated with antibiotics due to clinical or biochemical signs of infection. Patients who received a diagnosis of CL were extracted from the whole cohort, and they were compared to the counterparts who did not develop a CL. Hospital costs were defined as the charge of part of the hospitalization, from POD 1 until discharge. We did not include the costs of surgery because of the different costs of the 3 surgical procedures considered. Thus, hospital costs included the costs of antibiotics, nutritional support, postoperative imaging study, surgical or interventional postoperative procedure, and intensive care unit admission due to CL. The cost analysis was subdivided according to the 3 categories of CL identified by the ISGPS.4 The financial settlement system is regulated by the Italian Ministry of Health. All amounts were expressed in euros. Statistical analysis Data were analyzed using SPSS version 22.0 (IBM Corp, Armonk, NY). Normally distributed data are presented as mean with standard deviation (SD), and skewed distributed data are presented as median with interquartile range. The group differences were assessed using a Chi-square test or Fisher’s exact test for categorical variables. For non-normally distributed continuous variables the Mann-Whitney U test was applied. Patients were divided into 3 categories: CL grade A, B, and C according to the ISGPS definition.4 Clinical outcomes were calculated across all grades to show distribution of the outcomes between the CL groups. Statistical significance was determined by a P-value of <.05. Results The final population was 945 patients; 558 (58.4%), 311 (32.5%), and 87 (9.1%) patients underwent PD, DP, and TP, respectively. Considering the ISGPS definition and classification of CL, a total of 43 cases (4.5%) of CL were reported. Among the cases of CL, 10 (23.3%), 31 (72.1%), and 2 (4.6%) were registered as grade A, grade B, and grade C, respectively. Table 1 shows the demographic and
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
ARTICLE IN PRESS
JID: YMSY
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
3
Table 1 Demographic data. Study population n = 956
Age (years, median, IRQ) Sex (male) BMI (kg/m2 , IRQ) Diabetes ASA score III–IV CACI > 4 Jaundice Preoperative biliary drain Neoadjuvant therapy Pancreatic resection Pancreaticoduodenectomy Distal pancreatectomy Total pancreatectomy Surgical approach Open Laparoscopic Vascular resection Extended lymphadenectomy EBL (cc, median, IRQ) Operative time (min, median, IRQ) Length of stay (days, median, IRQ)
Total n (%)
No chyle leak 913 (95.5%)
Chyle leak 43 (4.5%)
P value
63 (54–71) 493 (51.6%) 24 (22–27) 185 (19.5%) 160 (16.7%) 360 (37.7%) 353 (36.9%) 301 (31.8%) 158 (16.5%)
64 (54–71) 473 (51.8%) 24 (22–27) 179 (19.7%) 155 (17%) 349 (38.2%) 339 (37.1%) 290 (32%) 149 (16.3%)
59 (50–69) 20 (46.5%) 23 (22–27) 6 (14%) 5 (11.6%) 16 (36.5%) 14 (32.5%) 11 (25.6%) 9 (20.9%)
.032∗ .300 .441 .237 .195 .310 .342 .238 .270 .947
558 (58.4) 311 (32.5) 87 (9.1%)
532(58.3%) 298 (32.6%) 83 (9.1%)
26 (60.5%) 13 (30.2%) 4 (9.3%)
855 (89.4%) 101 (10.6%) 121 (12.7%) 354 (37.1%) 30 0 (20 0–450) 375 (275–445) 9 (7–16)
815 (89.3%) 98 (10.7%) 115 (12.6%) 333 (36.6%) 30 0 (20 0–450) 375 (276–445) 9 (7–17)
40 (93%) 3 (7%) 6 (14%) 21 (48.8%) 30 0 (20 0–50 0) 390 (321–465) 14 (11–25)
.315
.469 .073 .895 .847 <.001∗
∗ Statistically significant. BMI: body mass index; ASA: American Society of Anesthesiology; CACI: Charlson Age-Comorbidity Index; EBL: estimated blood loss; IRQ: interquartile range.
Table 2 Pathological data. Study Population n = 956
Pathology PDAC IPMN Ampullary neoplasm Cholangiocarcinoma pNET Cystic lesion Other R+ resection Tumor size Lymph node harvest Positive lymph node harvest
Total n° (%)
No chyle leak 913 (95.5%)
Chyle leak 43 (4.5%)
445 (46.5%) 104 (10.9%) 78 (8.1%) 38 (4%) 142 (14.9%) 65 (6.8%) 84 (8.8%) 221 (23.6%) 25 (19–35) 36 (26–48) 1 (0–4)
427 (46.8%) 100 (11%) 75 (8.2%) 37 (4.1%) 133 (14.5%) 59 (6.4%) 82 (9%) 213 (23.8%) 25 (20–35) 36 (26–47) 1 (0–4)
18 (41.9%) 4 (9.3%) 3 (7%) 1 (2.3%) 9 (20.9%) 6 (13.9%) 2 (4.7%) 8 (18.6%) 25 (17–45) 33 (25–55) 1 (0–5)
P value .217
.280 .592 .589 .413
∗
Statistically significant. PDAC: pancreatic ductal adenocarcinoma; pNET: pancreatic neuroendocrine tumor; IPMN: intraductal papillary mucinous neoplasm.
surgical data. There was no association with preoperative and intraoperative variables, and only a younger age was associated with CL. The CL group had a longer, more statistically significant length of stay (P < .001). Table 2 shows pathological data; there were no similarities among CLs in the underlying pathology, the amount of lymph-node harvested, or metastatic in cases of malignant diseases. Table 3 reports the demographic and surgical data of CL cases. There was a clear prevalence of grade B cases, and the grade C was the rarest (4.6% of CL cases, = 2). A preoperative biliary drain was associated with grade B cases only. CL was reported mostly after PD with a clear prevalence of grade B cases. Higher estimated blood loss and a longer operative time were associated with a grade B CL. A higher lymph-node ratio was associated with grade B CL. CL alone occurred in 29 cases (67.4%). Considering all postoperative complications, grade B CL was associated more frequently with other complications and this was statistically significant for POPF, septic events, or major complications (P < .001, P = .008, and P = .004, respectively, Table 4). No deaths were recorded in this series. Of note, we report only 2 grade C cases. As presumed, a grade C CL was associated with a longer hospitalization; the mean ICU recovery was 7 days.
Economic analysis The economic analysis showed that CL burdened the costs of hospitalization with different severity according to the newly proposed ISGPS classification of CL. The average cost of hospitalization was indeed €2,806 ± €1,420 for grade A CL, €7,150 ± €4,791 for grade B CL, and €15,684 ± €473 for grade C CL (Fig. 1, P < .001). Discussion Different from POPF, postoperative pancreatic hemorrhage, and delayed gastric emptying, CL is an uncommon complication following pancreatic surgery. Recently, analyzing retrospectively a large series of 3,324 pancreatic resections, Strobel et al reported a 10% incidence of CL. The authors also report that preexisting diabetes, surgery duration >180 minutes, distal pancreatectomy, the concomitant presence of POPF, and resection for malignancy are all independent risk factors for CL.17 This latter finding is interesting yet not surprising; currently, the neoadjuvant therapy has a large diffusion, especially in pancreatic surgery. The tissue distortion caused by chemo- and/or radiotherapy might play a role
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
ARTICLE IN PRESS
JID: YMSY 4
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
Fig. 1. Cost analysis of chyle leak.
Table 3 Demographic and surgical data of CL cases. Total chyle leakage n = 43
Age (years, mean, SD) Sex (male) BMI (kg/m2 , SD) Diabetes ASA score III-IV CACI > 4 Jaundice Preoperative biliary drain Neoadjuvant therapy Pancreatic resection Pancreaticoduodenectomy Distal pancreatectomy Total pancreatectomy Surgical approach Open Laparoscopic Vascular resection Extended lymphadenectomy EBL (cc, median, IRQ) Operation time (min, median, IRQ)
Grade A10 (23.3%)
Grade B31 (72.1%)
Grade C2 (4.6%)
P value
56.3 (13.2) 3 (30%) 24.9 (4.5) 1 (10%) 1 (10%) 360 (37.7%) 1 (10%) 0 (0%) 1 (10%)
58.1 (12.8) 15 (48.4%) 24 (4.1) 5 (16.1%) 4 (12.9%) 349 (38.2%) 12 (42.9%) 11 (35.5%) 8 (25.8%)
67.5 (4.5) 2 (100%) 27.5 (2.1) 0 (0%) 0 (0%) 16 (36.5%) 0 (0%) 0 (0%) 0 (0%)
.447 .179 .314 .749 .845 .310 .098 .049∗ .428 .016∗
3 (30%) 7 (70%) 0 (0%)
22 (71%) 5 (16.1%) 4 (12.9%)
1 (50%) 1 (50%) 0 (0%)
8 (80%) 2 (20%) 1 (10%) 2 (20%) 215 (205) 295.5 (144.6)
30 (96.8%) 1 (3.2%) 4 (12.9%) 18 (58.1%) 403.2 (226.2) 40 0.7 (10 0.2)
2 (100%) 0 (0%) 1 (50%) 1 (50%) 300 (282.8) 382.5 (60.1)
.179
.313 .085 .022∗ .032∗
∗ Statistically significant. BMI: body mass index; ASA: American Society of Anesthesiology; CACI: Charlson Age-Comorbidity Index; EBL: estimated blood loss; IRQ: interquartile range.
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009
JID: YMSY
ARTICLE IN PRESS
[m5G;June 25, 2018;8:56]
S. Paiella et al. / Surgery 000 (2018) 1–5
5
Table 4 Postoperative complications associated with CL. Total chyle leakage n = 43
POPF Abdominal collection DGE Percutaneous drainage PPH Sepsis Clavien-Dindo score ≥ III Length of stay (days, median, IRQ) R+ resection Tumor size Lymph node harvest Positive lymph node harvest
Grade A10 (23.3%)
Grade B31 (72.1%)
Grade C2 (4.6%)
P value
0 3 0 0 0 0 0
(0%) (30%) (0%) (0%) (0%) (0%) (0%)
5 (16.1%) 18 (58.1%) 3 (9.7%) 3 (9.7%) 7 (22.6%) 9 (29%) 6 (19.4%)
1 2 1 0 0 2 2
(50%) (100%) (50%) (0%) (0%) (100%) (100%)
<.001∗ .121 .084 .536 .198 .008∗ .004∗
9.7 (2.2) 0 (0%) 50.8 (41.2) 24.5 (17.4) 0.5 (1.6)
19.8 (10) 8 (25.8%) 30.6 (20.5) 39.6 (15.9) 4.3 (7.2)
38 (12.7) 0 (0%) 35 (14.1) 27.5 (14.8) 0 (0)
<.001∗ .280 .201 .089 .030∗
∗ Statistically significant. DGE: delayed gastric emptying; POPF: postoperative pancreatic fistula; PPH: postoperative pancreatic hemorrhage.
in causing alterations in the structure of peripancreatic lymphatic vessels. Other studies suggest different predisposing factors, such as more extended lymphadenectomies,2 concomitant vascular resections,3 universal use of early enteral feeding,3,18 the manipulation of the para-aortic area and the retroperitoneal invasion,18 and the development of postoperative portal vein thrombosis.3 However, the results reported so far came from heterogeneous studies; hence no reliable conclusions could be made on the clinical impact of CL after pancreatic surgery. Finally, the ISGPS proposed a definition and a classification of CL. The present study is the first to apply this classification on a large single-center retrospective cohort of patients submitted to pancreatic surgery. In our experience, the most frequent grade was B. This is the result of our institutional treatment policy of CL, based on the early administration of TPN, rather than on the prescription of a low-fat diet. The use of TPN accelerates the resolution of the CL while the patient continues oral nutrition; this result is of paramount importance from a psychological standpoint. Our results confirm the validity of the ISGPS classification. In fact, the length of stay and the rates of major complications and septic events stratified consistently among the 3 grades. Even from an economic point of view, the new ISGPS classification depicted notable differences in the hospital costs among the 3 grades of CL, with a difference of €4,344 between grade A and grade B CL, €12,878 between grade A and grade C CL, and €8,534 between grade B and grade C CL. In our experience, the extended lymphadenectomy was not associated with a higher rate or a greater severity of CL. The analysis of the risk factors for CL was beyond the aims of this study. For the sake of completeness, we report a statistically significant association between CL and younger age only. A grade B CL was more frequently associated with preoperative biliary drain, DP, high estimated blood loss, and longer operative time. Conclusion In our experience, CL is a rare complication after pancreatic surgery, with a non-negligible impact on clinical and economic burden on the postoperative course. The newly proposed ISGPS classification identifies reliably and consistently the different clinical scenarios of CL and eliminates the previous heterogeneity of results in reporting CL after pancreatic surgery. References
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
A consensus statement by the international study group on pancreatic surgery (ISGPS). Surgery. 2014;156:591–600. Assumpcao L, Cameron JL, Wolfgang CL, Edil B, Choti MA, Herman JM, et al. Incidence and management of chyle leaks following pancreatic resection: A high volume single-center institutional experience. J Gastrointest Surg. 2008;12:1915–1923. Abu Hilal M, Layfield DM, Di Fabio F, Arregui-Fresneda I, Panagiotopoulou IG, Armstrong TH, et al. Postoperative chyle leak after major pancreatic resections in patients who receive enteral feed: Risk factors and management options. World J Surg. 2013;37:2918–2926. Besselink MG, van Rijssen LB, Bassi C, Dervenis C, Montorsi M, Adham M, et al. Definition and classification of chyle leak after pancreatic operation: A consensus statement by the International Study Group on pancreatic surgery. Surgery. 2017;161:365–372. Malleo G, Marchegiani G, Salvia R, Butturini G, Pederzoli P, Bassi C. Pancreaticoduodenectomy for pancreatic cancer: The Verona experience. Surg Today. 2011;41:463–470. Melotti G, Butturini G, Piccoli M, Casetti L, Bassi C, Mullineris B, et al. Laparoscopic distal pancreatectomy: Results on a consecutive series of 58 patients. Ann Surg. 2007;246:77–82. Butturini G, Damoli I, Crepaz L, Malleo G, Marchegiani G, Daskalaki D, et al. A prospective non-randomised single-center study comparing laparoscopic and robotic distal pancreatectomy. Surg Endosc. 2015;29:3163–3170. Callery MP, Pratt WB, Kent TS, Chaikof EL, Vollmer Jr CM. A prospectively validated clinical risk score accurately predicts pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg. 2013;216:1–14. McMillan MT, Malleo G, Bassi C, Allegrini V, Casetti L, Drebin JA, et al. Multicenter, prospective trial of selective drain management for pancreatoduodenectomy using risk stratification. Ann Surg. 2017;265:1209–1218. Bassi C, Molinari E, Malleo G, Crippa S, Butturini G, Salvia R, et al. Early versus late drain removal after standard pancreatic resections: Results of a prospective randomized trial. Ann Surg. 2010;252:207–214. Salvia R, Malleo G, Butturini G, Dal Molin M, Esposito A, Marchegiani G, et al. Perioperative management of patients undergoing pancreatic resection: Implementation of a care plan in a tertiary-care center. J Surg Oncol. 2013;107:51–57. 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. Chang CM, Yin WY, Wei CK, Wu CC, Su YC, Yu CH, et al. Correction: Adjusted age-adjusted Charlson comorbidity index score as a risk measure of perioperative mortality before cancer surgery. PLoS One. 2016;11. Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;161:584–591. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al. Postpancreatectomy hemorrhage (PPH): An International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery. 2007;142:20–25. Wente MN, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, et al. Delayed gastric emptying (DGE) after pancreatic surgery: A suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142:761–768. Strobel O, Brangs S, Hinz U, Pausch T, Huttner FJ, Diener MK, et al. Incidence, risk factors and clinical implications of chyle leak after pancreatic surgery. Br J Surg. 2017;104:108–117. Kuboki S, Shimizu H, Yoshidome H, Ohtsuka M, Kato A, Yoshitomi H, et al. Chylous ascites after hepatopancreatobiliary surgery. Br J Surg. 2013;100:522–527.
1. Tol JA, Gouma DJ, Bassi C, Dervenis C, Montorsi M, Adham M, et al. Definition of a standard lymphadenectomy in surgery for pancreatic ductal adenocarcinoma:
Please cite this article as: S. Paiella et al., Chyle leak after pancreatic surgery: validation of the International Study Group of Pancreatic Surgery classification, Surgery (2018), https://doi.org/10.1016/j.surg.2018.05.009