If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Postoperative pancreatic fistula (POPF) as well as postoperative biliary fistula (POBF) are considered the main source of postoperative morbidity and mortality after pancreatoduodenectomy (PD). However, little is known about the incidence and complications of combined POPF/POBF compared to isolated POPF or POBF.
Methods
This single-center study investigated retrospectively the incidence and postoperative outcome of combined POPF/POBF compared to isolated fistulas following PD in a tertiary German pancreatic center between 2009 and 2018.
Results
A total of 678 patients underwent PD for benign and malignant periampullary lesions. Combined fistulas occurred in 6%, isolated POPF in 16%, and isolated POBF in 2%. Pancreatic ductal adenocarcinoma and chronic pancreatitis had a protective effect on the occurrence of combined fistulas, whereas serous cystadenoma and pancreatic metastasis were risk factors. Morbidity (Grade C fistula, post-pancreatectomy hemorrhage, revisional surgery) and mortality was significantly higher in patients with combined fistulas than in those with isolated fistula. Moreover, the duration of ICU stay was longer.
Conclusions
A combined POPF/POBF is associated with a significant increase of morbidity and mortality compared to isolated fistulas after PD. Early surgical revision in these patients may improve the postoperative survival rate.
Introduction
Pancreatoduodenectomy (PD) is a well-established therapy to treat chronic pancreatitis, periampullary tumors, and other lesions of the pancreatic head. Furthermore, for malignant tumors multimodal therapy including R0 surgical resection offers a chance for cure. However, PD still carries a high risk of potentially life-threatening complications. Reasons for that are the technically demanding pancreatic and bile duct anastomoses, the corrosive secretions of both organs, and the topographical proximity to major blood vessels (e.g., major branches of the coeliac trunk or the superior mesenteric artery). Thus, although the surgical technique has improved and patients are preferably treated in highly specialized tertiary pancreatic centers, morbidity rates remain high (30–50%).
Besides the length of hospital stay and economic aspects, this is problematic especially for cancer patients, as it delays the beginning of an adjuvant therapy and worsens quality of life.
Postoperative organ fistula is a significant source of morbidity. The incidences of the postoperative pancreatic fistula (POPF) and the postoperative biliary fistula (POBF) after PD are 9–21% and 4–12%, respectively. Both fistulas are associated with sepsis, multi-organ failure, bleeding, and death. However, a leakage of the pancreaticoenteric anastomosis is probably the most feared complication, as it may cause life-threatening post-pancreatectomy hemorrhage (PPH).
Pancreatoduodenectomy (PD) and postoperative pancreatic fistula (POPF): a systematic review and analysis of the POPF-related mortality rate in 60,739 patients retrieved from the English literature published between 1990 and 2015.
It can be assumed that a combined fistula is associated with higher morbidity and mortality rates than an isolated POPF or POBF, however, this has – to the best of our knowledge – not yet been analyzed. Thus, the presented retrospective single-center study investigated the incidence, morbidity and mortality of combined pancreatic and biliary fistulas compared to isolated POPF or POBF after PD in a German tertiary pancreatic center. Potentially predisposing factors were additionally evaluated.
Methods
This retrospective single center study was performed at a German University hospital between 2009 and 2018. It was approved by the local Ethics committee of the Ruhr University Bochum (No.17-6251) and performed in accordance with the Helsinki Declaration as revised in 2013. Informed consent for surgery as well as for data collection and analysis was given by all patients.
Patients
We included all patients who underwent PD for resection of benign or malignant periampullary lesions (pylorus preserving Whipple procedure or classic Whipple operation). Patients who obtained any other pancreatic resection were excluded.
The following data were collected: age, gender, surgical procedure, underlying pathology, number of isolated POPF and POBF, combined POPF/POBF, PPH, length of hospital stay, length of ICU stay, need for re-surgery, 30- and 90-day mortality.
Surgical techniques
All procedures were conducted by 5 experienced pancreatic surgeons, who regularly performed more than 50 pancreatic resections per year, according to a standardized surgical technique. Pancreaticojejunostomy was carried out as end-to-side, duct-to-mucosa, double-layer anastomosis using interrupted polydioxanone (PDS) 5-0 sutures for the outer layer and interrupted polypropylene 5-0 sutures for the inner layer. Approximately 15 cm distal to the pancreaticojejunostomy, a single-layer end-to-side hepaticojejunostomy followed using interrupted PDS 5-0 sutures. Another 45–50 cm distal to the hepaticojejunostomy reconstruction was completed with antecolic duodenojejunostomy in a double-layer continuous PDS 4-0 suture technique as pylorus preserving Whipple procedure. In case of a classic Whipple procedure, a side-to-side Braun entero-enterostomy using double-layer continuous sutures with PDS 5–0 was added to gastrojejunostomy. In all cases two intraabdominal soft silicone drains were placed near the pancreaticojejunostomy and hepaticojejunostomy and derived separately in the left and right middle abdomen, respectively. The hepaticojejunostomy was splinted by T-tube in case of a tiny and thin-walled bile duct. All patients received 200 μg octreotide subcutaneous during induction of anesthesia.
Postoperative outcomes
POPF and PPH were defined according to the 2016 update of the International Study Group on Pancreatic Surgery (ISGPS).
The simultaneous appearance of amylase, lipase, and bilirubin (each of them at least 3-times above the corresponding serum concentration) was labelled as combined fistula independent of the affected drain. The detection of a bile admixture in the drain near the pancreaticojejunostomy was defined as a combined retrograde fistula (Fig. 1).
Figure 1Sketch, clarifying combined (a) and combined retrograde (b) fistulas after PD: green shows flow direction of bile, yellow – of pancreatic juice, HJ – hepatico-jejunostomy, PJ – pancreatico-jejunostomy, POPF - postoperative pancreatic fistula, POBF - postoperative biliary fistula, CF- combined fistula
Data were expressed as percentages, mean ± standard deviation, or median with minimum–maximum range as appropriate. All parametric variables of independent samples were analyzed using t-Tests. To compare more than 2 groups the ANOVA was undertaken. The adjacent post hoc analysis helped to explore differences between groups. Nonparametric data were explored by means of Mann–Whitney U-test. To show a relationship between two nominal variables, the Pearson's Chi square test was performed. If statistically significant differences were observed, an ensuing in-depth analysis was performed among the groups. A p-value of <0.05 was considered statistically significant. The analysis was performed with SPSS software (SPSS Inc., Chicago, Ill., USA).
Results
A total of 678 patients underwent open PD for benign and malignant periampullary lesions. An isolated POPF was observed in 16% (108/678), whereas 2% (16/678) suffered from an isolated POBF. A combined pancreatic-biliary fistula was seen in 6% (40/678). Patient demographics and the underlying diseases are presented in Table 1.
Table 1Patient demographics and underlying diseases
Malignancies represented most underlying diseases in all three groups (no fistula: p < 0.001; isolated fistula: 0.044; combined fistula: p = 0.040), however, malignancy itself was neither a risk factor nor protective to prevent the development of fistulas (all groups: p = 0.312; isolated vs. combined: p = 0.437).
A subgroup analysis revealed that the entities chronic pancreatitis and pancreatic ductal adenocarcinoma prevented the occurrence of combined fistulas (p = 0.049) and isolated fistulas (p < 0.001), respectively, whereas a serous cystadenoma (SCN) and pancreatic metastasis were risk factors for combined (p = 0.014) and isolated fistulas (p < 0.001) (Table 1).
Outcome parameters are presented in Table 2. The frequency of Grade C fistula, which is associated with severe morbidity and mortality, was significantly higher in patients with combined fistulas compared to isolated fistula (p = 0.006), whereas the less severe Grade A fistula were more frequently observed in patients with isolated fistula (p = 0.026). Furthermore, combined fistulas were more often complicated by PPH than isolated fistulas (p < 0.001).
Table 2Outcome parameters. ∗ Multiple entries possible
Isolated POPF
Isolated POBF
p-value
Combined fistula
All patients
p-value
(n = 108)
(n = 16)
(POPF vs. POBF)
(n = 40)
(n = 678)
(iso. vs. comb.)
Severity of fistula [%]
Grade A
29 (31/108)
19 (3/16)
0.406
10 (4/40)
6 (38/678)
0.026
Grade B
58 (62/108)
56 (9/16)
0.880
55 (22/40)
14 (93/678)
0.111
Grade C
13 (15/108)
25 (4/16)
0.206
35 (14/40)
5 (33/678)
0.006
Post-pancreatectomy hemorrhage [%]
6 (6/108)
6 (1/16)
0.999
25 (10/40)
2.5 (17/678)
< 0.001
CT-Drainage [%]
2 (2/108)
0
0.323
8 (3/40)
1 (5/678)
0.070
T- tube [%]
8 (9/108)
19 (3/16)
13 (5/40)
10 (65/678)
0.610
Revisional surgery∗ [%]
17 (18/108)
44 (7/16)
0.013
48 (19/40)
6.5 (44/678)
< 0.001
Completion-PE
72 (13/18)
0
–
53 (10/19)
3 (23/678)
0.199
Re-do-HJ
0
57 (4/7)
–
53 (10/19)
2 (14/678)
0.794
Re-do-PJ
11 (2/18)
0
–
0
0.3 (2/678)
–
Other
167 (3/18)
43 (3/7)
0.422
47 (9/19)
2 (15/678)
0.776
ICU stay (days) [mean ± SD]
4.2 ± 9.9
3.4 ± 5.2
0.753
10.3 ± 16.2
1.8 ± 3.2
0.013
Hospital stay (days) [mean ± SD]
29.5 ± 25.5
31.5 ± 17.1
0.762
35.6 ± 17.1
14 ± 6.8
0.354
Mortality [%]
30-days
2.8 (3/108)
0
0.807
7.5 (3/40)
2
0.136
90-days
4,6 (5/108)
6 (1/16)
0.522
15 (6/40)
2
0.031
PE – pancreatectomy, HJ – hepatico-jejunostomy, PJ – pancreatico-jejunostomy, ICU – intensive care unit, POPF – postoperative pancreatic fistula, POBF – postoperative biliary fistula.
Revisional surgery was performed significantly more often in patients suffering from isolated POBF or combined fistulas compared to isolated POPF (p = 0.013 and <0.001, respectively).
In 10 of 19 surgically revised patients with combined fistula a completion pancreatectomy with simultaneous re-do-hepaticojejunostomy was performed. In contrast to isolated POPF, no re-do-pancreaticojejunostomy was undertaken in this group. An erosive bleeding occurred twice in 2 patients after revisional surgery. Finally, 6 patients with combined fistula died after re-operation. There was no significant difference between all the groups concerning the revisional surgery related mortality.
Patients with combined fistulas resided at ICU significantly longer than patients with isolated fistula (p = 0.013). Moreover, mortality was significantly higher in this group, especially due to complicated POPF (p = 0.031). Fig. 2 demonstrates significantly high rate of secondary complications in combined fistulas compared to isolated POPF and POBF.
Figure 2Diagram demonstrates significantly high rate of secondary complications in combined fistulas compared to isolated POPF and POBF: FS-C – fistula severity grad C, PPH – postpancreatectomy hemorrhage, RS – revisional surgery, M−90 – 90-day mortality, I-POPF – isolated POPF, I–POBF – isolated POBF, CF – combined fistula
Despite the advances in surgical technique the pancreaticojejunostomy as well as the hepaticojejunostomy remain very challenging and high-risky procedure in PD. The POPF presents the most common and feared complication with incidences between 9 and 50%.
Pancreatoduodenectomy (PD) and postoperative pancreatic fistula (POPF): a systematic review and analysis of the POPF-related mortality rate in 60,739 patients retrieved from the English literature published between 1990 and 2015.
Systematic review and meta-analysis of 73 postoperative pancreatic fistula rates using the updated 2016 International Study Group Pancreatic Fistula definition in patients undergoing pancreatic resection with soft and hard pancreatic texture.
In this study a POPF occurred in 22% (16% isolated POPF and 6% combined fistulas). The reported incidence of biliary leakage is comparatively lower and varies between 4 and 12%.
In our study group the incidence of biliary leakage was 8% (2% isolated POBF and 6% combined fistulas), which is within the expected range compared to other specialized centers.
In 6% of our patients a combined fistula was observed. A concomitant appearance of both, POPF and POBF, is rarely mentioned by other authors and thus, comparative values are difficult to demonstrate. Fu et al. presented a combined fistulas rate of 7.7% in 532 patients after PD, however, further details were not given.
El Nakeeb et al. analyzed 555 patients with biliary complications after PD. Ten of these patients (1.8%) developed a biliary leakage concomitant to POPF, and four patients underwent re-surgery due to biliary peritonitis and associated POPF.
The etiology of combined fistulas remains unclear, whereas the risk factors contributing to insufficiency of the pancreaticojejunostomy or the hepaticojejunostomy are widely investigated. A tiny and thin-walled bile duct (<5 mm), bile infection, and compromised blood supply may lead to biliary leak.
Pancreatoduodenectomy (PD) and postoperative pancreatic fistula (POPF): a systematic review and analysis of the POPF-related mortality rate in 60,739 patients retrieved from the English literature published between 1990 and 2015.
In our study demographics did not influence the development of combined fistulas. However, chronic pancreatitis and pancreatic ductal adenocarcinoma prevented the occurrence of combined fistulas. Explanations might be that both entities are associated with a rather hard pancreatic tissue, as well as enlarged pancreatic and bile duct diameter. In contrast, pancreatic metastasis and SCN were risk factors for fistulas, as the tissue of the pancreas is soft, and the ducts are mostly small. Besides that, a reciprocal potentiation of bile and pancreatic juice may increase the severity of the fistula. It has been reported bile activates phospholipase A2 and it converts biliary lecithin to lysolecithin. The last ones demonstrate high cytotoxic activity. Moreover, bile provides an optimal alkaline pH at 8–8.5, which contributes to converting of trypsinogen to trypsin due to enteric enterokinase. Furthermore contaminated bile contains lipopolysaccharids, which potentiate the activity of pancreatic enzymes. These mechanisms demonstrate a high potentiating effect of bile on pancreatic juice, increasing proinflammatory and detrimental impact in case of combined fistulas.
In 25% of our patients, the combined fistulas were classified as retrograde fistulas. Here, a retrograde bile flow towards the pancreaticojejunostomy could be of impact.
In case of retrograde bile fistulas, the reconstructive technique could matter. Double loop reconstruction and other modifications, including a modified single loop reconstruction, could prevent the reflux of bile toward the pancreaticojejunostomy as recently described.
There are few RCTs and observational clinical studies comparing PG (pancreaticogastrostomy) vs PJ and showing no significant difference between these techniques concerning the rate of POBF and POPF. The occurrence of combined fistulas is not described in these studies explicitly. Nevertheless, PG might prevent a combined retrograde fistula, as well as a double loop or a modified single loop reconstruction. Probably the biliary reflux is one of the possible but not only cause of combined fistula particularly in case of concomitant but pure POPF and POBF. Splinting the hepaticojejunostomy by T-tube is used in patients with thin bile duct diameter or fragile bile duct wall. However, this technique does not prevent the biliary leakage. It solely aims to avoid re-surgery in case of insufficiency of the hepaticojejunostomy.
The analysis of the severity of fistulas in each group showed a significantly prevelance of Grade C fistulas in patients with combined insufficiency of the hepaticojejunostomy and the pancreaticojejunostomy. Along with that, Grade A fistulas ammounted the majority in patients with isolated POPF and POBF. It emphasizes an elevated morbidity of combined fistulas regardless of the mechanism of occurrence (retrogradely or separately). This is also reflected in the need for a longer ICU-stay, as well as in the increased 90-days mortality rate. Furthermore, our data correspond to the results of a study conducted by Jester et al., who investigated 924 patients after PD. A combined insufficiency of the hepaticojejunostomy and the pancreaticojejunostomy was seen in 31 patients (3%). This group had a significantly increased morbidity (58%), a longer median length of stay (14 days vs. 7–9 days), as well as a higher 90-day mortality (32% vs. 3.6%) compared to isolated POPF or no leak.
One of the most feared complications, the PPH, was observed in 17 patients in our study (10% of patients with fistulas). This corresponds to incidences reported by others (5–16%).
Detailed examination revealed that patients with combined fistulas were significantly more often affected by PPH than those with isolated fistula. The potentiating effect of bile leakage on PPH has already been reported.
The primary therapeutic approach in case of insufficiency of the pancreaticojejunostomy is rather conservative. Revisional surgery with completion pancreatectomy remains an ultima ratio (Fig. 3) and is associated with a high morbidity and mortality rate.
Certainly, the willingness for (early) surgery in patients with combined fistulas is higher to avert complications such as PPH, peritonitis, and sepsis, as well as to reduce the related mortality. That is why patients underwent revisional surgery more frequently than patients with isolated POPF in this study. However, it should be mentioned that some patients required repetitive re-surgeries (isolated and combined fistulas), which is associated with a significant increase of mortality (13–55%).
Figure 3CT-scan (a) and intraoperative images of combined fistula Grade C (b - POBF, c - POPF) as well as specimen after completion pancreatectomy (d): narrows show perihepatic ascites with free intraperitoneal air; A – insufficiency of pancreatico-jejunostomy, B – insufficiency of hepatico-jejunostomy, J-jejunum, L - liver, P - pancreas, S - spleen
This study demonstrated that a combined POPF/POBF fistula is associated with a significant increase of morbidity and mortality compared to isolated fistulas after PD. Chronic pancreatitis and pancreatic ductal adenocarcinoma prevent the occurrence of combined fistulas, whereas pancreatic metastasis and serous cystadenoma were risk factors. Early surgical revision in patients with combined fistulas may improve the postoperative survival rate.
There may be some possible limitations in this study. The first one is its retrospective design. The second limitation concerns to the limited number of patients in each group, particularly in the group of isolated POBF and combined fistulas. Third, it is difficult to differentiate between combined retrograde fistula and disruption of both anastomoses only basing on drain investigation, particularly in case of amylase/lipase and bilirubin elevation in the same drain. Fourth, there is a little number of prior studies, thus extensive comparison of results is difficult. Further research is needed to shed more light on the pathogenesis and prevention options of combined fistulas after PD.
Author contributions
Ilgar Aghalarov – acquisition, analysis and interpretation of data, drafting the paper, revising, final approving.
Elisabeth Beyer – acquisition, analysis and interpretation of data, revising, final approving.
Jennifer Niescery – analysis and interpretation of data, drafting the paper, revising, final approving.
Orlin Belyaev – analysis and interpretation of data, revising, final approving.
Waldemar Uhl – research design, analysis and interpretation of data, revising, final approving.
Torsten Herzog – research design, analysis and interpretation of data, revising, final approving.
Funding
No funding received.
Conflict of interest
None declared.
Acknowledgement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
References
He J.
Ahuja N.
Makary M.A.
Cameron J.L.
Eckhauser F.E.
Choti M.A.
et al.
2564 resected periampullary adenocarcinomas at a single institution: trends over three decades.
Pancreatoduodenectomy (PD) and postoperative pancreatic fistula (POPF): a systematic review and analysis of the POPF-related mortality rate in 60,739 patients retrieved from the English literature published between 1990 and 2015.
Systematic review and meta-analysis of 73 postoperative pancreatic fistula rates using the updated 2016 International Study Group Pancreatic Fistula definition in patients undergoing pancreatic resection with soft and hard pancreatic texture.
00051-5&id=gr1.jpg){kind=link}
00051-5&id=gr2.jpg){kind=link}
00051-5&id=gr3.jpg){kind=link}