Abstract
Background:
The short-term morbidity associated with post-operative pancreatic fistula (POPF) is well established, however data regarding the long-term impact are lacking. We aim to characterize long-term oncologic outcomes of POPF after pancreatic resection through a single institution, retrospective study of pancreatic resections performed for adenocarcinoma from 2009–2016.
Methods:
Kaplan-Meier survival analysis, logistic regression, and multivariate analysis (MVA) were used to evaluate impact of POPF on overall survival (OS), disease free survival (DFS), and receipt of adjuvant chemotherapy (AC).
Results:
767 patients were included. 82 (10.6%) developed grade B (n=67) or C (n=15) POPF. Grade C POPF resulted in decreased OS when compared to no POPF (20.22 vs 26.33 months, p=0.027) and to grade B POPF (20.22 vs. 26.87 months, p=0.049). POPF patients were less likely to receive AC than those without POPF (59.5% vs 74.9%, p=0.003) and grade C POPF were less likely to receive AC than all others (26.7% vs 74.2%, p=0.0001).
Discussion:
POPF patients are less likely to receive AC and more likely to have delay in time to AC. These factors are exacerbated in grade C POPF and likely contribute to decreased OS. These findings validate the clinical significance of the ISGPF definition of POPF.
Introduction
Post operative pancreatic fistula (POPF) remains one of the most challenging complications following pancreatic resection, occurring in up to one third of patients, despite a multitude of advances in both intra-operative technique and postoperative management (1–5). The International Study Group on Pancreatic Fistula (ISGPF) defined POPF in 2005 and updated the definition in 2016. The definition allows accurate comparison of occurrence rates across centers as well as the ability to study outcomes (1,6). Immediate postoperative implications of POPF are well established and include intra-abdominal abscess, perioperative hemorrhage, and increased length of stay, sepsis, and death. The ISGPF grading system has been validated in its ability to stratify patients with POPF in regard to these early outcomes (7).
Long-term implications associated with POPF are less clear. A limited number of studies have evaluated long-term outcomes associated with POPF with conflicting conclusions(8,9). Furthermore, the ISGPF grading criteria have not been validated concerning long-term outcomes. The oncologic impact of postoperative leak has been well described in colorectal, esophageal, and gastric cancer (10,11). Prior work has suggested that local inflammation associated with leaks for these cancers is a causative factor leading to recurrence, suggesting that POPF could also play a role in cancer recurrence.
The existing literature is inconclusive in regard to long-term oncologic impacts of pancreatic fistula. The aim of the present study is to characterize the impact of POPF on long-term outcomes following pancreatic resection for adenocarcinoma.
Methods
Study Design and Data Collection
All patients who underwent pancreatic resection for an adenocarcinoma (peri-ampullary and pancreatic) from March 2009 to December 2016 at University of Pittsburgh Medical Center were identified retrospectively from an IRB approved institutional database (PRO #19060135). Patients who had undergone robotic or open Whipple, and robotic, laparoscopic, or open distal pancreatectomy were included. Peri-ampullary adenocarcinomas were included in addition to pancreatic ductal adenocarcinoma (PDAC) in order to increase the sample size for the study. Parameters collected in retrospective review of patient data included patient demographics, tumor characteristics, pathologic staging data (node status, lymphovascular and perineural invasion, grade), treatment variables (procedure performed, R status, neoadjuvant or adjuvant chemotherapy, time to adjuvant therapy), post operative complications (POPF, delayed gastric emptying (DGE), wound infection, venous thromboembolism (VTE), and length of stay), and outcomes (survival, recurrence status). Patients were then subdivided into groups based on presence of POPF and further subdivided into 2016 ISGPF grade of POPF (no POPF, grade B POPF, grade C POPF). Patients who died within 90 days postoperatively were excluded. This exclusion allowed examination of long-term outcomes rather than acute post-operative mortality. The primary outcome measured was overall survival (OS). OS was calculated from the date of oncologic operation to date of last known follow up or death. For those patients who received neoadjuvant therapy, we elected not to use date of neoadjuvant therapy initiation for OS calculation to avoid confounding the OS analysis. Secondary outcomes included disease free survival (DFS), receipt of adjuvant chemotherapy (AC) and time to receipt of AC. DFS was calculated from the date of operation to date of documented recurrence, last known follow up, or death.
Statistical Analysis
For the analysis of differences between groups in continuous variables with a normal distribution, unpaired t test was used while the Wilcoxon rank-sum was used for variables without normal distribution. For the analysis of differences between groups in categorical variables, Chi-squared or Fisher’s exact test was used according to data size. Indicator variables for POPF grades were created with group of no POPF as a reference. Univariate and multivariate logistic regression modeling was conducted to predict factors of receiving AC and results presented as odds ratios (OR) with the 95% confidence intervals (CI). Kaplan-Meier method was used to show differences in OS, DFS, and differences in time to AC between study groups. Differences between median OS, median DFS, and median time to AC were tested using the log-rank test. Cox proportional hazard regression models were performed for predicting OS and presented as hazard ratios (HR) with the 95% CI. Similarly, Cox proportional hazard regression models were used to predict time to AC. A sub-group analysis in patients with PDAC was performed to show the effect of POPF grade, with related interaction term, on OS and DFS. All multivariate analyses were performed in a backward stepwise analysis. All tests were of 2-sided nature and statistical significance assigned when p≤0.05. Statistical analyses were performed using the Intercooled Stata 13.0 statistical software package, College Station, TX: StataCorp LP.
Results
Patient Cohort Characteristics
819 patients who underwent pancreatic resection for adenocarcinoma from 2009 to 2016 were identified. Because of the aim to evaluate long-term outcomes of POPF, 52 patients were excluded due to mortality within 90 days from operation. Of the excluded patients, 6 developed POPF (11.5%). Of the 6 that developed POPF, 3 were grade B and 3 were grade C. 4 of the excluded patients who developed POPF underwent open whipple and 2 underwent robotic whipple. 90 day mortality was secondary to POPF in 4 of the 6 excluded patients that developed POPF. The remaining causes of death in the excluded patients were; sepsis with etiology other than POPF (cholangitis, pneumonia, clostridium difficile colitis, bile leak, wound infection) (n=11), gastroduodenal artery bleed (n=3), bowel ischemia (n=3), organ failure (hepatic, renal, pulmonary) (n=9), gastrojejunostomy leak (n=1), intrahepatic biliary-arterial fistula (n=1), myocardial infarction/cardiac arrest (n=3), stroke (n=2), operative death (n=1). The remaining (n=14) were not determined or attributed to failure to thrive. An analysis of the entire cohort of patients is presented in Supplemental Table 1. Of the 767 patients included in the study, the average age was 67.2 +/− 10.6 years and there were 364 (47.5%) females. POPF occurred in 82 patients (10.7%), of those 67 were grade B POPF (81.7%) and 15 were grade C POPF (18.3%). Of those 82 patients that developed POPF, 55 (67%) were in patients with PDAC versus 27 (33%) in patients with other peri-ampullary adenocarcinomas.
When demographics of the POPF and no POPF groups were compared, there were no significant differences (Table 1). Comparison of pre-operative variables including American Society of Anesthesiologists (ASA) class, Charlson Comorbidity Index (CCI), albumin level, and Ca 19–9 level demonstrated no differences.
Table 1.
Patient Demographics, Tumor Characteristics, and Treatment Variables
| All-cohort N=767 | No POPF N=685 | POPF N=82 | p | |
|---|---|---|---|---|
| Age at diagnosis (years) | 67.2±10.6 | 67.4±10.5 | 65.2±10.8 | 0.078 |
| Age > 65 years | 471 (61.4) | 424 (61.9) | 47 (57.3) | 0.421 |
| Sex (F) | 364 (47.5) | 331 (48.3) | 33 (40.2) | 0.166 |
| ASA | 0.590 | |||
| 2 | 97 (12.7) | 84 (12.3) | 13 (16.0) | |
| 4 | 59 (7.7) | 51 (7.4) | 8 (9.9) | |
| CCI age adjusted | 4.94±1.69 | 4.98±1.70 | 4.66±1.58 | 0.106 |
| Pre-operative albumin (g/dL) | 3.56±0.55 | 3.55±0.56 | 3.64±0.52 | 0.211 |
| Ca19–9 % change | −73 (−89, −31.2) | −74 (−89, 31.4) | −38 (−80, −27) | 0.301 |
| T-size (cm) | 2.93±1.48 | 2.91±1.47 | 3.09±1.59 | 0.289 |
| PDAC | 578 (75) | 523 (76) | 55 (67) | 0.077 |
| T stage | 0.139 | |||
| 1 | 55 (12.7) | 46 (6.7) | 9 (11) | |
| 4 | 39 (5.1) | 32 (4.7) | 7 (8.5) | |
| N stage | 0.148 | |||
| 2 | 11 (1.4) | 8 (1.2) | 3 (3.7) | |
| Lymphovascular Invasion | 548 (72.4) | 492 (72.8) | 56 (69.1) | 0.488 |
| Perineural Invasion | 591 (77.6) | 534 (78.5) | 57 (69.5) | 0.064 |
| Grade | 0.796 | |||
| 2 | 511 (68.4) | 458 (68.8) | 53 (65.4) | |
| 4 | 4 (0.6) | 4 (.6) | 0 (0) | |
| Neoadjuvant Chemotherapy | 331 (43.5) | 316 (46.4) | 15 (18.8) | 0.0001* |
| Operative Time (minutes) | 398.5 ± 137.0 | 392.7±131.6 | 447.9±169.7 | 0.001* |
| EBL (cc) | 300 (150, 600) | 300 (150, 550) | 400 (150, 800) | 0.068 |
| EBL >=500 cc | 243 (33.4) | 210 (32.1) | 33 (45.2) | 0.024* |
| Vein resection | 148 (19.3) | 141 (20.6) | 7 (8.5) | 0.009* |
| Whipple | 642 (83.7) | 577 (84.2) | 65 (79.3) | 0.250 |
| Robotic | 374 (48.8) | 347 (50.7) | 27 (32.9) | 0.002* |
| R1 resection | 116 (15.2) | 106 (15.5) | 10 (12.2) | 0.425 |
| Post-op transfusion | 167 (22.9) | 151 (23.1) | 16 (21.6) | 0.781 |
| Wound infection | 157 (20.5) | 131 (19.5) | 26 (31.7) | 0.008* |
| VTE | 41 (5.5) | 35 (5.3) | 6 (7.8) | 0.362 |
| DGE | 150 (20.2) | 127 (19.1) | 23 (29.5) | 0.031* |
| Adjuvant chemotherapy | 537 (73.3) | 490 (74.9) | 47 (59.5) | 0.003* |
| Time to adjuvant chemotherapy (days) | 62 (60, 65) | 61 (58, 63) | 75 (68, 84) | 0.002* |
Comparison of pathologic variables including T and N stage, pathology (PDAC versus non-PDAC), grade, and lymphovascular and perineural invasion demonstrated no significant differences (Table 1). 578 patients (75%) were resected for PDAC and 189 for peri-ampullary malignancy. Peri-ampullary malignancies included those diagnosed as duodenal adenocarcinoma, cholangiocarcinoma, and ampullary adenocarcinoma or mucinous adenocarcinoma. Average tumor size for the cohort was 2.93 cm.
There were several significant differences when comparing intraoperative variables between the two patient cohorts. Patients who developed POPF had a longer median operative time, were more likely to have an estimated blood loss >500cc, were more likely to have had an open operation, and were less likely to have required a vein resection (Table 1).
Differences in post-operative variables between the groups included higher rates of wound infection and delayed gastric emptying (DGE), as well as longer median length of hospital stay in the POPF group. There were no differences in VTE or post-operative transfusion requirement.
Kaplan-Meier Survival Analysis
There were no differences in OS or DFS between patients with or without POPF. Further analysis demonstrated significantly decreased median OS in patients with grade C POPF versus no POPF (20.22 vs 26.33 months, p=0.027) and in patients with grade C POPF versus grade B POPF (20.22 vs 26.87 months, p=0.049) as depicted in Figure 1. With regard to DFS, Kaplan-Meier analysis demonstrated significantly decreased median DFS in patients with grade C POPF versus no POPF (8.0 vs 15.73 mo p=0.031) and in patients with grade C POPF versus grade B POPF (8.0 vs 19.77 p=0.014) as depicted in Figure 2.
Figure 1. Kaplan-Meier overall survival estimates by POPF status.
OS curves for no POPF, Grade B, and Grade C POPF are depicted. Significant differences between the Grade C and Grade B curves (p=0.049) and between the Grade C and no POPF curves (p=0.027) are identified. There is no difference between the Grade B and no POPF curves (p=0.690).
Figure 2. Kaplan-Meier disease free survival estimates by POPF status.
DFS curves for no POPF, Grade B, and Grade C POPF are depicted. Analysis demonstrates a significant difference between the Grade C and Grade B curves (p=0.014) and between the Grade C and no POPF curves (p=0.031). There is no difference between the Grade B and no POPF curves (p=0.207).
Multivariable Survival Analysis
On multivariate analysis, grade C POPF was an independent predictor of OS (HR 2.04 CI 1.11,3.74 p=0.020). Receipt of AC could not be fit into the cox regression model due to co-linearity with POPF (Table 2). When PDAC and PDAC+POPF were included as variables in the multivariate analysis, the combined PDAC+POPF variable was an independent predictor of OS (HR 1.52 CI 1.09, 2.12 p=0.013). Further analysis with a combined variable of PDAC+Grade C POPF demonstrated it as an independent predictor of OS. (HR 2.92 CI 1.43,5.95 p=0.003).
Table 2.
Multivariate cox regression analysis for predictors of overall survival
| Variable | HR | 95% CI | p |
|---|---|---|---|
| ASA | 1.32 | 1.06–1.63 | 0.012 |
| Preoperative albumin | 0.78 | 0.66–0.93 | 0.006 |
| T stage | 1.39 | 1.18–1.65 | 0.00 |
| N stage | 1.41 | 1.14–1.74 | 0.001 |
| R1 resection | 1.80 | 1.40–2.31 | 0.00 |
| Vein resection | 1.39 | 1.09–1.78 | 0.008 |
| Grace B POPF | 1.17 | 0.84–1.63 | 0.360 |
| Grade C POPF | 2.04 | 1.11–3.74 | 0.022 |
| PDAC Pathology | 1.19 | 0.92–1.52 | 0.170 |
Adjuvant Chemotherapy Analysis
POPF patients were less likely to receive AC than those without POPF (59.5% vs 74.9% p=0.003) and POPF patients had longer median time to AC than those without (75 vs 61 days, p=0.002). Grade C POPF patients were less likely to receive AC than all others (26.7% vs 74.2% p=0.0001). Of the 15 patients with grade C POPF, 14 had indications for AC (PDAC pathology, node positivity, lymphovascular or perineural invasion). On logistic regression modeling grade C POPF was independently associated with failure to receive AC (OR 0.19 CI 0.05, 0.65 p=0.009)
Pathology Subset Analysis
Rates of POPF development in specific cohorts were further analyzed. In comparison of pathology cohorts, 9.52% of patients with PDAC developed POPF while 14.29% of patients with non-PDAC pathology developed POPF (p=0.065). In the cohort of patients with PDAC, POPF was associated with significantly worse OS (21.3 vs 24.93 mo p=0.020). (Figure 3A). Within the PDAC group, there was no significant difference in DFS in those with POPF and those without (10.37 vs 15.03 mo p=0.175). As reported above, when PDAC and PDAC+POPF were introduced as variables in the MVA, PDAC+POPF and PDAC+ Grade C POPF were independent predictors of OS.
Figure 3. Kaplan-Meier survival estimates by POPF status and pathology.
(A) OS for patients with PDAC pathology, separated into those with and without POPF. There is a significant difference between these curves (p=0.02). (B) DFS for patients with PDAC pathology, separated into those with and without POPF. There is not a significant difference between these curves (p=0.175).
Operation Subset Analysis
In comparison of operative groups, the rate of POPF in those undergoing whipple was 10.1% versus 13.6% in distal pancreatectomy (p=0.267). There were no significant differences in OS or DFS in this subset (Supplemental Figure 1).
Neoadjuvant Subset Analysis
15 of the 331 (4.5%) patients who received neoadjuvant therapy developed POPF, which was significantly lower that than the 15.4% in those that did not receive neoadjuvant (p<0.0001). Of those 15 patients that developed POPF, 4 of them were grade C. Hence, 1.2% of patients receiving neoadjuvant chemotherapy developed grade C POPF versus 2.5% of patients that did not receive neoadjuvant (p=0.2925). Neoadjuvant chemotherapy consisted of gemcitabine or FOLFIRINOX based regimens in 327 of the 331 patients. Of patients who received neoadjuvant therapy, 80.1% received chemotherapy alone. The remaining 19.9% of patients who received neoadjuvant therapy, received chemotherapy and radiation. In comparing these two groups, the rate of POPF in those who received chemotherapy only was 4.33% versus 6.06% of those who received chemotherapy and radiation (p=0.511).
Of the 82 patients that developed POPF, 17 (20.7%) of them had received neoadjuvant therapy and 65 (79.3%) did not. There was not a significant difference in median OS between these two subsets (14.76 vs 26.73 p=0.156)
Discussion
Despite significant improvements in pancreatic surgery over the last several decades, POPF remains a common and morbid complication. While the short-term impact of POPF has been well established, data on the long-term outcomes associated with POPF are lacking. In the current study, we evaluate a large, single institution cohort of pancreatectomy patients, demonstrating that grade C POPF is associated with worse long term survival. Additionally, patients with POPF were significantly less likely to receive adjuvant chemotherapy.
The present study is the largest cohort studied in regard to long-term outcomes after POPF and the first since the updated ISGPF definition of POPF. Previously published data on long term outcomes following POPF have been conflicting in nature. Nagai et al described long-term outcomes associated with POPF in a retrospective review of 184 patients undergoing resection of PDAC with curative intent. The 2005 ISGPF criteria were used to define POPF. Of 184 patients, POPF occurred in 51 (27.7%). In their cohort, POPF was significantly associated with peritoneal recurrence and was an independent risk factor with hazard ratio (HR) 3.974. Additionally, POPF was the strongest independent risk factor for peritoneal recurrence among those examined including AC, tumor size and differentiation, local invasion, and resection margin. Local recurrence was not reported. This increased risk of peritoneal recurrence did not apparently impact survival, as there was no difference in OS in those with POPF. Those with POPF did have a shorter DFS that did not reach statistical significance (8). A more recent 2013 study by Assifi et al conversely concluded no impact of POPF on tumor recurrence and long-term outcomes. The group performed a retrospective analysis of 221 patients that underwent pancreaticoduodenectomy with curative intent for PDAC. Of 221 patients, 23 had POPF (10.4%) with grade A n=9, grade B n= 13, grade C n=1 (using the 2005 ISGPF definition). Of note, the group did not routinely measure drain amylase and POPF was diagnosed clinically. Peritoneal recurrence in those patients with POPF was 13% vs. 10% in those without POPF. There were similarly no differences in local recurrence, 22% in the POPF group versus 21% in those without POPF. Additionally, they demonstrated no significant difference in DFS or OS. On multivariate analysis, POPF was not an independent risk factor for recurrence free survival (RFS) or OS. Vascular invasion, tumor stage and differentiation, and adjuvant radiation were identified as independent risk factors for both RFS and OS. Based on these data, Assifi et al concluded that tumor recurrence is independent of POPF in patients undergoing pancreaticoduodenectomy for PDAC. (9)
We found no differences in OS or DFS when comparing patients with POPF to those without, however when stratified by grade of POPF we detected significant differences with grade C POPF having worse OS and DFS than grade B POPF and those without POPF. This stepwise decrease in OS and DFS between POPF B and C validates the updated ISGPF definition of POPF with regard to long-term outcomes and can provide qualitative prognostic information.
Patients with grade C POPF were less likely to receive AC and were more likely to have a delay in receipt of AC. This likely contributes to the decrease in median DFS and OS. On multivariate analysis, grade C POPF was an independent predictor of decreased median OS, however AC could not be fit into the model due to co-linearity with POPF, indicating a likely association between the two. It has been well established that AC prolongs OS and DFS. The CONKO-001 trial published in 2013 initially demonstrated increased OS and DFS in patients with PDAC with 6 months of gemcitabine therapy compared with observation alone, which solidified AC as standard of care (12). The study found a median DFS of 13.4 months versus 6.7 months, and median OS of 22.8 months versus 20.2 months in the gemcitabine versus observation groups. These data support the notion that the benefit of AC may be the lone etiology of survival differences between POPF and no POPF groups.
Our subset analysis within PDAC patients alone again demonstrated decreased OS in those with POPF. There was a trend toward decreased DFS in patients with POPF that did not reach significance. This analysis was underpowered due to small sample size of PDAC patients that developed POPF, and particularly grade C POPF (only 10 patients).
Our subset analysis of those patients who received neoadjuvant chemotherapy demonstrated a significantly lower rate of POPF formation in comparison to those that did not. This suggests a protective effect of neoadjuvant therapy and lends further support to neoadjuvant chemotherapy for all patients with PDAC in addition to the demonstrated benefits in regard to R0 resection rates and survival (13,14). While promising, this effect is less apparent when grade C POPF are examined. There is no difference in rate of grade C POPF between patients that did and did not receive neoadjuvant chemotherapy. Treatment effect could become apparent in a larger cohort of patients.
This study is limited by its retrospective design. The prominent limitation of the study is the low sample size in regard to number of POPF within the cohort, with only 15 patients developing a grade C POPF. This necessitated an analysis of all peri-ampullary adenocarcinoma and PDAC pathologies as opposed to PDAC alone. This additionally limited the ability to control for variables during multivariate analysis, such as receipt of AC, which was particularly limiting given the strong association between POPF and receipt of AC. A larger, multi-institutional cohort would be beneficial to further investigate the long-term nuance of POPF, particularly within PDAC patients.
The exclusion of patients with 90-day postoperative mortality could certainly impact conclusions regarding overall POPF outcomes, however the goal of the current manuscript was to characterize long-term outcomes associated with POPF.
An additional limitation of the study is the difficulty of ascertaining decisions regarding AC in a retrospective fashion. This may have influenced the data regarding AC in patients with POPF, in that AC may not have been indicated or pursued, rather than not administered secondary to POPF.
Despite these limitations, we can conclude that POPF does in fact impact long-term outcomes following pancreatic resection for peri-ampullary adenocarcinomas. Patients with grade C POPF are less likely to receive AC, and have decreased OS and DFS compared to grade B POPF and all others. This provides valuable prognostic information for this group of patients. A larger cohort of patients is required to further analyze grade B and C POPF while controlling for receipt of AC and to study POPF within PDAC patients alone.
These findings validate the updated ISGPF criteria for POPF in regard to long-term outcomes. This validation serves to provide confidence in the way that POPF are studied and reported in the literature.
These findings necessitate further basic science and translational study regarding the biologic mechanisms that may be driving these outcomes. We postulate that the inflammatory environment associated with POPF may contribute to recurrence, as has been demonstrated in other cancers. Future studies will aim to investigate the role of POPF associated inflammation in carcinogenesis and recurrence.
Supplementary Material
Acknowledgements
Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number 5U54GM104942-04 (BB). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Submitted for review as an original article. Work presented in the form of a poster presentation at the Society of Surgical Oncology Annual Meeting 2019.
References
- 1.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. Surg (United States) [Internet].2017;161(3):584–91. Available from: 10.1016/j.surg.2016.11.014 [DOI] [PubMed] [Google Scholar]
- 2.Søreide K, Healey AJ, Mole DJ, Parks RW. Pre-, peri- and post-operative factors for the development of pancreatic fistula after pancreatic surgery. Hpb [Internet].2019; Available from: https://linkinghub.elsevier.com/retrieve/pii/S1365182X1930591X [DOI] [PubMed] [Google Scholar]
- 3.Kawaida H, Kono H, Hosomura N, Amemiya H, Itakura J, Fujii H, et al. Surgical techniques and postoperative management to prevent postoperative pancreatic fistula after pancreatic surgery. World J Gastroenterol. 2019;25(28):3722–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Søreide K, Labori KJ. Risk factors and preventive strategies for post-operative pancreatic fistula after pancreatic surgery: a comprehensive review. Scand J Gastroenterol. 2016;51(10):1147–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Berger AC, Howard TJ, Kennedy EP, Sauter PK, Bower-Cherry M, Dutkevitch S, et al. Does Type of Pancreaticojejunostomy after Pancreaticoduodenectomy Decrease Rate of Pancreatic Fistula? A Randomized, Prospective, Dual-Institution Trial. J Am Coll Surg [Internet].2009;208(5):738–47. Available from: 10.1016/j.jamcollsurg.2008.12.031 [DOI] [PubMed] [Google Scholar]
- 6.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(1):8–13. [DOI] [PubMed] [Google Scholar]
- 7.Pulvirenti A, Marchegiani G, Pea A, Allegrini V, Esposito A, Casetti L, et al. Clinical Implications of the 2016 International Study Group on Pancreatic Surgery Definition and Grading of Postoperative Pancreatic Fistula on 775 Consecutive Pancreatic Resections. Ann Surg. 2017;268(6). [DOI] [PubMed] [Google Scholar]
- 8.Nagai S, Fujii T, Kodera Y, Kanda M, Sahin TT, Kanzaki A, et al. Recurrence pattern and prognosis of pancreatic cancer after pancreatic fistula. Ann Surg Oncol. 2011;18(8):2329–37. [DOI] [PubMed] [Google Scholar]
- 9.Assifi MM, Zhang S, Leiby BE, Pequignot EC, Xia B, Rosato E, et al. Tumor recurrence is independent of pancreatic fistula in patients after pancreaticoduodenectomy for pancreatic ductal adenocarcinoma. J Am Coll Surg [Internet].2013;217(4):621–7. Available from: 10.1016/j.jamcollsurg.2013.05.014 [DOI] [PubMed] [Google Scholar]
- 10.Law WL, Choi HK, Lee YM, Ho JWC, Seto CL. Anastomotic leakage is associated with poor long-term outcome in patients after curative colorectal resection for malignancy. J Gastrointest Surg. 2007;11(1):8–15. [DOI] [PubMed] [Google Scholar]
- 11.Andreou A, Biebl M, Dadras M, Struecker B, Sauer IM, Thuss-Patience PC, et al. Anastomotic leak predicts diminished long-term survival after resection for gastric and esophageal cancer. Surg (United States) [Internet].2016;160(1):191–203. Available from: 10.1016/j.surg.2016.02.020 [DOI] [PubMed] [Google Scholar]
- 12.Oettle H, Neuhaus P, Hochhaus A, Hartmann JT, Gellert K, Ridwelski K, et al. Adjuvant chemotherapy with gemcitabine and long-term outcomes among patients with resected pancreatic cancer: The CONKO-001 randomized trial. JAMA - J Am Med Assoc. 2013;310(14):1473–81. [DOI] [PubMed] [Google Scholar]
- 13.Rangarajan K, Pucher P, Armstrong T, Bateman A, Hamady Z. Systemic neoadjuvant chemotherapy in modern pancreatic cancer treatment: a systematic review and meta-analysis. Ann R Coll Surg Engl. 2019;101(7):453–462. doi: 10.1308/rcsann.2019.0060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Janssen QP, Buettner S, Suker M, et al. Neoadjuvant FOLFIRINOX in Patients With Borderline Resectable Pancreatic Cancer: A Systematic Review and Patient-Level Meta-Analysis. J Natl Cancer Inst. 2019;111(8):782–794. doi: 10.1093/jnci/djz073 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.