Original article

Risk of pancreatic fistula after enucleation of pancreatic

tumours
O. Strobel1 , A. Cherrez1 , U. Hinz1 , P. Mayer2 , J. Kaiser1 , S. Fritz1 , L. Schneider1 , M. Klauss2 ,
M. W. Büchler1 and T. Hackert1
Departments of 1 Surgery and 2 Diagnostic and Interventional Radiology, Heidelberg University Hospital, Heidelberg, Germany
Correspondence to: Dr O. Strobel, Department of Surgery, Heidelberg University Hospital, Im Neuenheimer Feld 110, 69120 Heidelberg, Germany
(e-mail: oliver.strobel@med.uni-heidelberg.de)

Background: Enucleation is used increasingly for small pancreatic tumours. Data on perioperative
outcome after pancreatic enucleation, especially regarding the significance and risk factors associated
with postoperative pancreatic fistula (POPF), are limited. This study aimed to assess risk-dependent
perioperative outcome after pancreatic enucleation, with a focus on POPF.
Methods: Patients undergoing enucleation for pancreatic lesions between October 2001 and February
2014 were identified from a prospective database. A detailed analysis of morbidity was performed. Risk
factors for POPF were assessed by univariable and multivariable analyses.
Results: Of 166 enucleations, 94 (56⋅6 per cent) were performed for cystic and 72 (43⋅4 per cent) for
solid lesions. Morbidity was observed in 91 patients (54⋅8 per cent). Severe complications occurred in
30 patients (18⋅1 per cent), and one patient (0⋅6 per cent) died. Reoperation was necessary in nine
patients (5⋅4 per cent). POPF was the main determinant of outcome and occurred in 68 patients (41⋅0
per cent): grade A POPF, 34 (20⋅5 per cent); grade B, ten (6⋅0 per cent); and grade C, 24 (14⋅5 per cent).
Risk factors independently associated with POPF were: cystic tumour, localization in the pancreatic tail,
history of pancreatitis and cardiac co-morbidity. Only cystic morphology was independently associated
with clinically relevant POPF (grade B or C), occurring after enucleation in 25 (27 per cent) of 94 patients
with cystic tumours versus nine (13 per cent) of 72 patients with solid tumours. Tumour size and distance
to the main duct were not associated with risk of POPF.
Conclusion: Enucleation is a safe procedure in appropriately selected patients with a low rate of severe
complications. POPF is the main determinant of outcome and is more frequent after the enucleation of
cystic lesions.
Presented to the 132nd Congress of the German Society of Surgeons, Munich, Germany, April 2015, and the 11th
Congress of the European–African Hepato-Pancreato-Biliary Association, Manchester, UK, April 2015

Paper accepted 1 April 2015

Published online 24 June 2015 in Wiley Online Library (www.bjs.co.uk). DOI: 10.1002/bjs.9843

Introduction context of high postoperative morbidity, decision-making

The mortality rate following pancreatic resections has been
reduced to around 3 per cent, but the morbidity rate
is still as high as 50 per cent, even in recent random-
ized clinical trials and in high-volume centres1 – 4 . Post-
operative pancreatic fistula (POPF) is the commonest and
most important complication after pancreatic resection,
being associated with, and triggering, most other major
complications5 .
With the broader use of cross-sectional imaging and
endoscopic ultrasonography, small pancreatic tumours
are increasingly detected as incidental findings6 – 8 . In the
regarding resection of these frequently benign and often
asymptomatic lesions is difficult. As a less invasive and
parenchyma-sparing procedure, enucleation is used
increasingly for small and presumed benign pancre-
atic lesions9 – 11 . Despite being less invasive than formal
resections, morbidity and, especially, POPF rates after
enucleation are still high, varying from 20 to 67 per
cent11 – 14 . Although the incidence of POPF and risk
factors associated with its development after formal pan-
creatic resections have been studied extensively3,15 – 17 ,
few studies, often with small numbers of patients, have

© 2015 BJS Society Ltd BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
Pancreatic fistula after enucleation of pancreatic tumours
addressed the risk of POPF following enucleation12 – 14 .
In contrast to resections, little is known about the clinical
significance of POPF after enucleation.
The aim of the present study was to assess the incidence
and significance of POPF, as well as risk factors for POPF,
after enucleation based on perioperative outcomes in a
large single-centre series.
Methods
As approved by the local ethics committee, perioperative
data of all patients undergoing pancreatic surgery at the
Department of General Surgery, Heidelberg University
Hospital, from October 2001 were collected in a specifi-
cally developed electronic database. Patients who had enu-
cleation for a pancreatic lesion between October 2001 and
February 2014 were identified, and their data analysed with
respect to demographic and preoperative parameters, sur-
gical and pathological findings, and perioperative outcome.
Patients who underwent concomitant surgical procedures
involving other organs were included if it was considered
that these could not affect outcomes specific to enucle-
ation.
Preoperative diagnostic investigation included
cross-sectional imaging by contrast-enhanced CT, MRI
or endoscopic ultrasonography with tissue confirmation.
In the majority of patients the indication for surgical
resection (versus observation) was based on the diagnosis of
a solid tumour or clinical symptoms. With growing expe-
rience and the detection of malignancy in Sendai-negative
branch-duct intraductal papillary mucinous neoplasms
(BD-IPMNs)18 , indications for surgical intervention were
extended to incidental smaller cystic lesions when a muci-
nous cystic neoplasm was suspected. Taking into account
patient age and co-morbidity, the indication for surgery in
suspected BD-IPMN was based on a size of at least 2 cm
or an increase in cyst size during observation.
The final decision to perform enucleation rather than for-
mal resection was based on intraoperative findings involv-
ing surgical exploration with inspection, palpation and
complete mobilization of the pancreas as necessary. In addi-
tion to preoperative imaging, intraoperative ultrasonog-
raphy was used to determine the exact location of the
lesion with respect to the main pancreatic duct and to
exclude multifocal lesions. Enucleation was attempted in
the absence of signs of malignancy or of direct involvement
of the main duct.
Enucleations were performed using careful microsurgi-
cal, mainly blunt, separation of the usually well defined
lesion and the lobuli of the pancreatic parenchyma.
© 2015 BJS Society Ltd
Published by John Wiley & Sons Ltd
1259
Vascular and ductal structures were divided after clo-
sure by clip ligature or monofilament sutures. Bipolar
coagulation was used only for superficial structures and
avoided in the vicinity of the main duct. Other technical
sealing devices were not used except for the laparoscopic
procedures. If a complete enucleation was not possible
without compromising the main pancreatic duct, the
operation was converted to a formal resection. Frozen
section of enucleated lesions was performed routinely, and
when there were doubts about the benign character of the
enucleated lesion the procedure was also converted to a
formal resection. In the database, converted procedures
were entered as the resulting formal resections and were
therefore not included in the present study. Suture of the
pancreatic parenchyma, coverage of the enucleation site
with a serosal patch, use of sealants, and intraoperative and
postoperative administration of octreotide were performed
at the surgeon’s discretion. One or two silicone drains were
routinely placed at the enucleation site.
Volume and appearance of drain outflow were recorded
daily. Drains were routinely removed on postoperative day
2 or 3, unless drain amylase, lipase or bilirubin levels were
raised. POPF was defined according to the International
Study Group on Pancreatic Fistula (ISGPF)5 . In patients
with POPF, intra-abdominal drains were kept in place and
removed later depending on the clinical course.
Postoperative CT was used liberally if there was any sign
of infection. When peripancreatic fluid collections were
identified, CT-guided drainage was attempted and patients
were treated with antibiotics. Repeat surgery was used only
when this strategy appeared unsuccessful.
Multiple preoperative, intraoperative and postoperative
parameters were extracted either from the prospective
database or from electronic and archived patient files.
Parameters collected in the prospective database included
patient age at operation, sex, American Society of Anes-
thesiologists (ASA) fitness grade, body mass index, type
of surgery, duration of the operation, histopathologi-
cal tumour entity, postoperative morbidity, POPF, need
for interventional therapy (CT-guided interventions,
endoscopic retrograde pancreatography) or reoperation,
postoperative hospital stay, and in-hospital and 30-day
mortality.
To exclude incomplete recording of clinically
non-relevant POPF in patients operated on before the
introduction of the ISGPF definition in 20055 , POPF was
determined based on a combined assessment of informa-
tion from the prospective pancreatic database and of data
points for amylase and lipase drain fluid collected prospec-
tively in the institutional laboratory data information
system.
www.bjs.co.uk BJS 2015; 102: 1258–1266
1260 O. Strobel, A. Cherrez, U. Hinz, P. Mayer, J. Kaiser, S. Fritz et al.

Table 1 Demographic and preoperative data Table 3 Perioperative outcome following pancreatic enucleation
No. of patients* (n = 166) No. of patients* (n = 166)

Age (years)† 58⋅2 (48⋅5–67⋅7) Hospital stay (days)† 8⋅5 (7–13)

Sex ratio (M : F) 43 : 123 Readmission 18 (10⋅8)
Body mass index (kg/m2 )† 24⋅8 (22⋅1–27⋅9) Overall morbidity 91 (54⋅8)
ASA fitness grade Clavien–Dindo grade21
I 11 (6⋅6) 0 (no complication) 75 (45⋅2)
II 110 (66⋅3) I 26 (15⋅7)
III 42 (25⋅3) II 35 (21⋅1)
Missing 3 (1⋅8) IIIa 21 (12⋅7)
Cardiac co-morbidity 41 of 165 (24⋅8) IIIb 7 (4⋅2)
Arterial hypertension 62 of 165 (37⋅6) IV 1 (0⋅6)
Diabetes mellitus 15 of 165 (9⋅1) V (death) 1 (0⋅6)
Symptomatic 100 of 163 (61⋅3) POPF
History of acute pancreatitis 16 (9⋅6) No 98 (59⋅0)
Radiological data‡ Grade A 34 (20⋅5)
Grade B 10 (6⋅0)
Cystic morphology 90 (66⋅2)
Grade C 24 (14⋅5)
Size of lesion (mm)† 19 (15–25)
Postpancreatectomy haemorrhage‡ 4 (2⋅4)
Minimum distance to main duct (mm)† 2⋅2 (1⋅1–4⋅9)
Delayed gastric empyting 8 (4⋅8)
*With percentages in parentheses unless indicated otherwise; †values are Acute pancreatitis 2 (1⋅2)
median (i.q.r.). ‡Included only if imaging results still available Bile leakage 5 (3⋅0)
electronically (30 missing). ASA, American Society of Anesthesiologists. Haematoma 4 (2⋅4)
Wound infection 13 (7⋅8)
Table 2 Surgical and pathological data CT without intervention 12 (7⋅2)
Interventional therapy
No. of patients* (n = 166)
CT-guided drainage 18 (10⋅8)
Surgical data ERCP 13 (7⋅8)
Tumour location Reoperation 9 (5⋅4)
Head and uncinate process 82 (49⋅4)
*With percentages in parentheses unless indicated otherwise; †values are
Body 53 (31⋅9)
median (i.q.r.). ‡All four progressed to grade C. POPF, postoperative pan-
Tail 21 (12⋅7)
creatic fistula; ERCP, endoscopic retrograde cholangiopancreatography.
Several sites 10 (6⋅0)
No. of enucleations
1 150 (90⋅4) Additional parameters extracted from patient files
2 14 (8⋅4) included cardiocirculatory and renal co-morbidities,
3 2 (1⋅2)
pre-existing diabetes mellitus, preoperative symptoms
Surgical approach
Open 163 (98⋅2) attributed to the tumour and history of acute pancre-
Laparoscopic 3 (1⋅8) atitis. Intraoperative parameters were exact location of
Coverage (with serosal patch) 64 (39⋅6) the lesion with respect to surgical anatomy, technique of
Use of sealants 15 (9⋅0)
enucleation, assessment of additional procedures, coverage
Intraoperative octreotide 58 of 165 (35⋅2)
Concomitant procedure of enucleation site, octreotide treatment and blood loss
Cholecystectomy 92 (55⋅4) estimated by the anaesthetist. Postoperative parameters
Adhesiolysis 45 (27⋅1) were tumour size as determined by the pathologist and a
Other‡ 38 (22⋅9)
Duration of surgery (min)† 138 (100–161)
detailed assessment of postoperative outcome.
Blood loss (ml)† 200 (100–300) Severity of POPF, delayed gastric emptying and post-
Surgical pathology data pancreatectomy haemorrhage were classified according to
Morphology the ISGPF5 and International Study Group of Pancre-
Cystic 94 (56⋅6)
Solid 72 (43⋅4)
atic Surgery (ISGPS)19,20 definitions. For patients who
Tumour entity had surgery before the introduction of these definitions,
Pancreatic neuroendocrine neoplasia 60 (36⋅1) information was extracted from patient files. All complica-
IPMN 64 (38⋅6) tions including postoperative acute pancreatitis, bile leak-
Benign mucinous cystic neoplasm 4 (2⋅4)
Serous cystic neoplasm 21 (12⋅7)
age, haematoma, wound infection, urinary tract infection,
Other entity 17 (10⋅2) pulmonary complications and rare complications were clas-
Tumour size (mm)† 13 (9–20) sified retrospectively according to Clavien–Dindo21 .
*With percentages in parentheses unless indicated otherwise; †values are Radiological data were recorded for all patients where
median (i.q.r.). ‡For details see Table S1 (supporting information). IPMN, preoperative cross-sectional imaging (CT and/or MRI)
intraductal papillary mucinous neoplasia. was still available electronically. Radiological parameters

© 2015 BJS Society Ltd www.bjs.co.uk BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
Pancreatic fistula after enucleation of pancreatic tumours
40 Only POPF grade A
Only POPF grade B
Only POPF grade C
POPF grade A and other complication
30 POPF grade B and other complication
POPF grade C and other complication
No. of patients
Only other complication
20
10
0 2
I II IIIa IIIb IV V No POPF
Grade of complications
a Contribution of POPF to overall morbidity
Impact of postoperative pancreatic fistula (POPF) on perioperative outcome: a overall morbidity (classified according to
Fig. 1
Clavien–Dindo21 ); b postoperative hospital stay. P < 0⋅001 (Kruskal–Wallis test); *P < 0⋅050 (Dunn’s multiple comparison test)
included tumour size (measured in the plane of maxi-
mum tumour extent), cystic morphology and minimal
distance from the tumour to the main pancreatic duct. The
parameters were determined using all available images
and orientations by two experienced radiologists with
the caliper tool within the picture archiving and com-
munication system (PACS) (GE Healthcare, Wauwatosa,
Wisconsin, USA).
Patients with missing data were omitted from the univari-
able and multivariable analyses.
Statistical analysis
SAS® software release 9.4 (SAS Institute, Cary, North
Carolina, USA) was used. Quantitative parameters are
expressed as median (i.q.r.) unless indicated otherwise. The
non-parametric Mann–Whitney U test was used to com-
pare quantitative parameters between groups. Categori-
cal parameters are presented as frequencies and compared
between enucleations with and without POPF using the
χ2 test, if appropriate, or Fisher’s exact test. Univariable
and multivariable logistic regression analyses were per-
formed to determine parameters associated with the risk of
developing POPF (any grade) and clinically relevant POPF
(grades B and C). Parameters with a P value below the
20 per cent level in univariable analysis were included in
the multivariable analysis. For the comparison of postop-
erative stay between patients without POPF and POPF
of different grades, Prism® version 5.04 (GraphPad Soft-
ware, San Diego, California, USA) and the Kruskal–Wallis
© 2015 BJS Society Ltd
Published by John Wiley & Sons Ltd
1261
∗
128
∗
64
Postop. hospital stay (days)
∗
32
16
8
Grade A Grade B Grade C
b Effect of POPF on postoperative hospital stay
test with Dunn’s multiple comparison test were used.
Two-sided P values were computed, and a difference was
considered statistically significant at P ≤ 0⋅050.
Results
Of 4329 pancreatic resections performed between October
2001 and February 2014, a total of 170 patients (3⋅9 per
cent) underwent enucleation. Four patients who had enu-
cleation as part of complex surgical procedures and experi-
enced morbidity that could not be attributed to the enucle-
ation itself were excluded from this study. Characteristics of
the 166 patients who underwent enucleation for a pancre-
atic lesion are summarized in Table 1. There was a predomi-
nance of women. Most patients had little co-morbidity and
were classified as ASA grade II. The pancreatic lesions were
symptomatic in most patients.
An overview of surgical data is provided in Table 2. Of
note, almost half of the lesions were located in the head
of the pancreas or the uncinate process, and would have
required a pancreatic head resection as the alternative sur-
gical procedure. About one-third of lesions were located
in the pancreatic body, requiring as an alternative either a
left resection associated with extensive parenchymal loss
or a segmental resection. When enucleation was accompa-
nied by a concomitant procedure, the most common was
cholecystectomy.
The majority of enucleated lesions were cystic (Table 2).
Complete excision was achieved in all lesions treated by
enucleation. Final pathology confirmed the non-invasive
www.bjs.co.uk BJS 2015; 102: 1258–1266
1262 O. Strobel, A. Cherrez, U. Hinz, P. Mayer, J. Kaiser, S. Fritz et al.

Table 4 Univariable analysis of variables associated with postoperative pancreatic fistula

POPF (any grade) POPF grade B or C

No. of patients No. of events P‡ No. of events P‡

Demographic/preoperative
Age (years) <50 46 22 0⋅312 9 0⋅929
50–70 93 38 20
≥70 27 8 5
Sex M 43 14 0⋅193 7 0⋅428
F 123 54 27
Body mass index (kg/m2 ) <18⋅5 3 2 0⋅443 0 0⋅467
18⋅5 to < 25 83 30 14
25 to < 30 56 24 15
≥30 24 12 5
ASA fitness grade* I 11 8 0⋅042 2 0⋅436
II 110 46 26
III 42 13 6
Cardiac co-morbidity* Yes 41 13 0⋅154 6 0⋅276
No 124 55 28
Diabetes mellitus* Yes 15 3 0⋅080 1 0⋅311
No 150 65 33
Symptomatic* Yes 100 41 0⋅868 19 0⋅798
No 63 25 13
History of acute pancreatitis Yes 16 11 0⋅017 5 0⋅325
No 150 57 29
Radiological†
Cystic morphology Yes 90 46 0⋅071 22 0⋅348
No 46 16 8
Size of lesion (mm) ≤19 70 28 0⋅128 15 0⋅781
>19 64 34 15
Minimum distance to main duct (mm) ≤ 2⋅2 67 33 0⋅405 14 0⋅510
>2⋅2 62 26 16
Surgical
Tumour location Head/uncinate 82 31 0⋅081 20 0⋅429
Body 53 18 8
Tail 21 13 3
Several sites 10 6 3
No. of enucleations 1 150 59 0⋅191 31 0⋅857
2 or 3 16 9 3
Coverage (serosal patch) Yes 64 28 0⋅563 16 0⋅253
No 102 40 18
Intraoperative octreotide* Yes 58 21 0⋅336 12 0⋅984
No 107 47 22
Concomitant procedure
Cholecystectomy Yes 92 33 0⋅137 18 0⋅744
No 74 35 16
Adhesiolysis Yes 45 20 0⋅578 11 0⋅440
No 121 48 23
Other Yes 38 14 0⋅556 7 0⋅720
No 128 54 27
Duration of surgery (min) <138 82 37 0⋅282 17 0⋅937
≥138 84 31 17
Blood loss (ml)* ≤ 200 89 38 0⋅204 21 0⋅412
>200 56 18 10
Surgical pathology
Morphology Cystic 94 49 0⋅001 25 0⋅026
Solid 72 19 9
Tumour entity pNEN 60 18 <0⋅001 8 0⋅192
IPMN 64 31 16
Benign MCN 4 3 1
SCN 21 14 7
Other entity 17 2 2

© 2015 BJS Society Ltd www.bjs.co.uk BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
Pancreatic fistula after enucleation of pancreatic tumours 1263

Table 4 Continued

POPF (any grade) POPF grade B or C

No. of patients No. of events P‡ No. of events P‡

Tumour size (mm)* ≤10 50 18 0⋅695 9 0⋅363

10–20 78 34 19
>20 37 15 5

*Data missing for some patients. †Included only if imaging results still available electronically (30 missing); data for lesion size and distance to pancreatic
duct missing for an additional two and seven patients respectively. POPF, postoperative pancreatic fistula; ASA, American Society of Anesthesiologists;
pNEN, pancreatic neuroendocrine neoplasia; IPMN, intraductal papillary mucinous neoplasia; MCN, mucinous cystic neoplasm; SCN, serous cystic
neoplasm. ‡χ2 test or Fisher’s exact test.

Table 5 Multivariable logistic regression analysis of variables associated with postoperative pancreatic fistula in 160 patients

POPF (any grade) POPF grade B or C

Hazard ratio P Hazard ratio P

Tumour entity
Benign MCN/SCN versus pNEN/other 7⋅76 (2⋅77, 23⋅95) <0⋅001
IPMN versus pNEN/other 1⋅93 (1⋅32, 6⋅65) 0⋅009
Localization (tail versus other) 4⋅45 (1⋅50, 14⋅33) 0⋅009
History of acute pancreatitis (yes versus no) 3⋅12 (0⋅98, 11⋅07) 0⋅061
Cardiac co-morbidity (yes versus no) 0⋅34 (0⋅13, 0⋅81) 0⋅018
Cystic morphology (yes versus no) 2⋅53 (1⋅13, 6⋅12) 0⋅030
Not included
Localization, body 0⋅905
Localization, several sites 0⋅825
Cholecystectomy 0⋅755
Male sex 0⋅520
ASA fitness grade (III versus < III) 0⋅421
Cystic morphology 0⋅329
Diabetes 0⋅239
Enucleations (1 versus 2/3) 0⋅237
IPMN 0⋅789
MCN/SCN 0⋅432

Values in parentheses are 95 per cent c.i. Six patients were excluded from the multivariable analysis owing to missing values. POPF, postoperative
pancreatic fistula; MCN, mucinous cystic neoplasm; SCN, serous cystic neoplasm; pNEN, pancreatic neuroendocrine neoplasia; IPMN, intraductal
papillary mucinous neoplasia; ASA, American Society of Anesthesiologists.

nature of all cystic tumours. Of 64 IPMNs, 53 had Postoperative pancreatic fistula

low-grade dysplasia, eight had moderate dysplasia, three
had high-grade dysplasia and none was invasive. All muci-
nous cystic neoplasms and serous cystic neoplasms were
benign.
Perioperative outcome
The main outcome data are shown in Table 3. The overall
morbidity rate, including events that were not clinically rel-
evant (such as POPF grade A or delayed gastric emptying
grade A), was 54⋅8 per cent (91 patients). The majority (61,
36⋅7 per cent) had minor complications (Clavien–Dindo
grade I or II21 ). Severe complications (grade III or above)
were observed in 30 patients (18⋅1 per cent), including one
death.
POPF occurred in 68 patients (41⋅0 per cent) and was
the most common complication. In half of these patients,
POPF was not clinically relevant (grade A). The rate of
clinically relevant POPF was 20⋅5 per cent (34 of 166
patients). Owing to an aggressive approach with early inter-
ventional management, clinically relevant POPF was most
often grade C. In contrast, the reoperation rate was only
5⋅4 per cent (9 patients), and only four patients (2⋅4 per
cent) had reoperation for complications associated with
POPF. Of the remaining five patients, three were reop-
erated for management of a bile duct lesion after enucle-
ation in the head of the pancreas, one for leakage from an
aberrant bile duct after cholecystectomy, and one for evac-
uation of a subcutaneous haematoma after concomitant

© 2015 BJS Society Ltd www.bjs.co.uk BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
1264 O. Strobel, A. Cherrez, U. Hinz, P. Mayer, J. Kaiser, S. Fritz et al.

No POPF POPF grade B Risk factors for postoperative pancreatic fistula

POPF grade A POPF grade C
The results of univariable analyses of parameters associ-
100
ated with the risk of POPF are summarized in Table 4.
90 Low ASA fitness grade, an episode of acute pancreatitis
80 in the patient’s history, cystic morphology of the lesion
70 by histopathology, and cystic tumour entities were signif-
icantly associated with an increased risk of POPF (any
% of patients

60
grade). In contrast, only cystic tumour entity was identi-
50
fied as a risk factor for clinically relevant POPF (grade B or
40 C). In multivariable analysis, cystic tumour entities, local-
30 ization in the tail of the pancreas, and cardiac co-morbidity
20 were identified as independent risk factors for POPF (any
grade); a history of acute pancreatitis showed a clear trend.
10
Only cystic morphology remained an independent risk fac-
0
Head/uncinate Body Tail Several sites tor for clinically relevant POPF (grade B or C) (Table 5).
Two parameters that might influence surgical
Frequency and grade of postoperative pancreatic fistula
Fig. 2 decision-making, owing to their presumed association
(POPF) in relation to the enucleation site(s) with POPF risk, size of the lesion and minimum distance
of the lesion to the main pancreatic duct, were not associ-
incisional hernia repair. In contrast to formal pancreatic ated with POPF. An increased risk for POPF of any grade
resections, POPF after pancreatic enucleation was associ- was seen following enucleations located in the pancreatic
ated with postpancreatectomy haemorrhage only rarely (4 tail, although this did not translate to clinically relevant
patients). POPF because most POPFs from left-sided enucleations
The contribution of POPF to overall morbidity after were clinically non-relevant (Fig. 2). Neither radiological
pancreatic enucleation is shown in Fig. 1a. Most patients nor pathological tumour size was associated with POPF
with morbidity after enucleation had POPF as their only or (Table 4). Similarly, the minimum distance of the lesion
a contributing complication, and only a few patients with to the main pancreatic duct was not associated with risk
clinically significant complications had no POPF. This is of POPF (of any grade, or for grade B or C), as tested
further demonstrated by the significant impact of POPF by entering the distance as a continuous variable or using
on postoperative hospital stay, which increased with POPF cut-off values. An example of such a lesion in direct contact
grade (Fig. 1b). with the main pancreatic duct is shown in Fig. 3.

1 cm

a Direct contact of hypervascular lesion with b Intact pancreatic main duct after enucleation
main duct

Fig. 3a Preoperative contrast-enhanced CT scan (arterial phase) in axial orientation showing a hypervascular lesion (pancreatic
neuroendocrine tumour) in the pancreatic body (arrow) in direct vicinity to the main pancreatic duct (arrowheads); distance between
lesion and main duct, 0 mm. b Operative images showing the enucleated lesion and an intact pancreatic duct after enucleation

© 2015 BJS Society Ltd www.bjs.co.uk BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
Pancreatic fistula after enucleation of pancreatic tumours 1265

Discussion of any coagulation or sealing device near the main duct. In

both studies12,13 that reported proximity to the main duct
POPF was the major determinant of perioperative outcome
as a risk factor, the rates of laparoscopic enucleation, where
after pancreatic enucleation. Cystic tumour morphology
sealing devices are frequently used, were higher (13–30
was the main risk factor for POPF. In the context of the
per cent) than the rate here. Univariable analysis also failed
high morbidity rate associated with anatomical pancreatic
to identify any association of coverage with a serosal patch
resection, pancreatic enucleation is a procedure with a
or administration of somatostatin analogues with POPF,
reduced rate of severe complications.
although in a recent randomized clinical trial24 a novel
Although the overall morbidity rate of 54⋅8 per cent may
somatostatin analogue markedly reduced clinical POPF
at first appear high, it compares favourably with morbidity
after formal resection and appears to be worthy of inves-
rates reported in previous series of enucleations, ranging
tigation in the context of enucleation. Unlike other series,
between 36⋅5 per cent without and 67⋅2 per cent with the
the present patient cohort contained a high proportion of
use of a comprehensive classification system that includes
cystic lesions, and these tumour entities were identified as
minor complications12,14 . The major complication rate of
risk factors for POPF of any grade and clinically relevant
18⋅1 per cent is much lower than that reported for for-
fistulas. The only other study25 to assess the results of enu-
mal resections in recent studies (30–50 per cent)1,3,4 . The
cleations for cystic lesions evaluated parenchyma-sparing
analysis of overall morbidity and hospital stay in the con-
resections specifically for IPMNs, including 44 enucle-
text of POPF clearly defined POPF as the major determi-
ations, and reported an overall POPF rate after these
nant of perioperative outcome after enucleation. Because
procedures of 55 per cent and a clinically relevant POPF
the overall surgical trauma following enucleation is con-
rate of 17 per cent, similar to the present observations.
siderably less than after pancreatoduodenectomy and distal
A weakness of the present study was the inability to
pancreatectomy, it might be assumed that the develop-
identify and include patients in whom an enucleation was
ment of POPF would have a larger relative impact on the
planned but who had conversion to a formal resection
postoperative course and recovery than after formal resec-
based on intraoperative evaluation. Such intention-to-treat
tions, but this was not the case. Haemorrhage and death
data would be valuable for preoperative decision-making.
after enucleation were uncommon events compared with
Another weakness of the study is that a part of the analysis
these outcomes after formal resection (postpancreatectomy
was based on retrospective data collection. This resulted
haemorrhage, 3–8 per cent; mortality, 2–3 per cent)1,3,4 .
in missing values, especially with regard to preoperative
There is usually no need to divide any major vessel during
radiological imaging, and impacts on the strength of the
enucleation, thus reducing the risk of erosion and bleeding
analysis to identify parameters based on preoperative imag-
from vessel stumps if a leak should occur. More aggres-
ing as predictors for POPF after enucleation.
sive endoscopic and interventional radiological manage-
Case–control studies and randomized trials for compara-
ment than often reported may also be relevant13 . The latter
tive assessment of perioperative and long-term outcome of
approach may also explain the low reoperation rate. An
enucleation versus formal resection now seem logical steps.
early and aggressive interventional approach to undrained
Given the increased risk of POPF in patients with cystic
fluid collections resulted in an upgrading of POPF from
tumours, such studies should include stratification and sub-
grade B to grade C, but in so doing probably reduced the
group analyses with respect to solid versus cystic tumour
need for reoperation22,23 .
morphology.
In previous studies that analysed risk factors for the
development of POPF, distance of the lesion of 2 mm or
less12 or 3 mm or less13 , cardiac insufficiency (New York Disclosure
Heart Association grade II or III) and duration of surgery
of 180 min or more were identified as independent risk The authors declare no conflict of interest.
factors for POPF. The present study, however, did not
confirm distance as a risk factor. Rather, the present results References
show that even tumours in the direct vicinity of the main
1 Correa-Gallego C, Brennan MF, D’Angelica M, Fong Y,
duct could be enucleated with comparable safety. This is Dematteo RP, Kingham TP et al. Operative drainage
of special interest for lesions in the pancreatic neck and following pancreatic resection: analysis of 1122 patients
body, where the pancreas is thin and even small tumours resected over 5 years at a single institution. Ann Surg 2013;
are very close to the main duct. One possible explanation 258: 1051–1058.
for the difference may be technique. These enucleations 2 Strobel O, Hinz U, Gluth A, Hank T, Hackert T, Bergmann
were undertaken as open procedures and avoided the use F et al. Pancreatic adenocarcinoma: number of positive nodes

© 2015 BJS Society Ltd www.bjs.co.uk BJS 2015; 102: 1258–1266

Published by John Wiley & Sons Ltd
1266
allows to distinguish several N categories. Ann Surg 2015;
261: 961–969.
3 Diener MK, Seiler CM, Rossion I, Kleeff J, Glanemann M,
Butturini G et al. Efficacy of stapler versus hand-sewn closure
after distal pancreatectomy (DISPACT): a randomised,
controlled multicentre trial. Lancet 2011; 377: 1514–1522.
4 Van Buren G II, Bloomston M, Hughes SJ, Winter J,
Behrman SW, Zyromski NJ et al. A randomized prospective
multicenter trial of pancreaticoduodenectomy with and
without routine intraperitoneal drainage. Ann Surg 2014;
259: 605–612.
5 Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C,
Izbicki J et al.; International Study Group on Pancreatic
Fistula. Postoperative pancreatic fistula: an international
study group (ISGPF) definition. Surgery 2005; 138: 8–13.
6 Crippa S, Partelli S, Zamboni G, Scarpa A, Tamburrino D,
Bassi C et al. Incidental diagnosis as prognostic factor in
different tumor stages of nonfunctioning pancreatic
endocrine tumors. Surgery 2014; 155: 145–153.
7 Robinson SM, Scott J, Oppong KW, White SA. What to do
for the incidental pancreatic cystic lesion? Surg Oncol 2014;
23: 117–125.
8 Sachs T, Pratt WB, Callery MP, Vollmer CM Jr. The
incidental asymptomatic pancreatic lesion: nuisance or
threat? J Gastrointest Surg 2009; 13: 405–415.
9 Briggs CD, Mann CD, Irving GR, Neal CP, Peterson M,
Cameron IC et al. Systematic review of minimally
invasive pancreatic resection. J Gastrointest Surg 2009; 13:
1129–1137.
10 Cauley CE, Pitt HA, Ziegler KM, Nakeeb A, Schmidt CM,
Zyromski NJ et al. Pancreatic enucleation: improved
outcomes compared to resection. J Gastrointest Surg 2012;
16: 1347–1353.
11 Hackert T, Hinz U, Fritz S, Strobel O, Schneider L, Hartwig
W et al. Enucleation in pancreatic surgery: indications,
technique, and outcome compared to standard pancreatic
resections. Langenbecks Arch Surg 2011; 396: 1197–1203.
12 Brient C, Regenet N, Sulpice L, Brunaud L,
Mucci-Hennekine S, Carrere N et al. Risk factors for
postoperative pancreatic fistulization subsequent to
enucleation. J Gastrointest Surg 2012; 16: 1883–1887.
13 Heeger K, Falconi M, Partelli S, Waldmann J, Crippa S,
Fendrich V et al. Increased rate of clinically relevant
pancreatic fistula after deep enucleation of small pancreatic
tumors. Langenbecks Arch Surg 2014; 399: 315–321.
14 Zhang T, Xu J, Wang T, Liao Q, Dai M, Zhao Y.
Enucleation of pancreatic lesions: indications, outcomes, and
Supporting information
Additional supporting information may be found in the online version of this article:
Table S1 Concomitant surgical procedures (Word document)
© 2015 BJS Society Ltd
Published by John Wiley & Sons Ltd
O. Strobel, A. Cherrez, U. Hinz, P. Mayer, J. Kaiser, S. Fritz et al.
risk factors for clinical pancreatic fistula. J Gastrointest Surg
2013; 17: 2099–2104.
15 Nathan H, Cameron JL, Goodwin CR, Seth AK,
Edil BH, Wolfgang CL et al. Risk factors for pancreatic
leak after distal pancreatectomy. Ann Surg 2009; 250:
277–281.
16 Gaujoux S, Cortes A, Couvelard A, Noullet S, Clavel L,
Rebours V et al. Fatty pancreas and increased body mass
index are risk factors of pancreatic fistula after
pancreaticoduodenectomy. Surgery 2010; 148: 15–23.
17 Kawai M, Kondo S, Yamaue H, Wada K, Sano K, Motoi F
et al. Predictive risk factors for clinically relevant pancreatic
fistula analyzed in 1239 patients with
pancreaticoduodenectomy: multicenter data collection as a
project study of pancreatic surgery by the Japanese Society of
Hepato-Biliary-Pancreatic Surgery. J Hepatobiliary Pancreat
Sci 2011; 18: 601–608.
18 Fritz S, Klauss M, Bergmann F, Hackert T, Hartwig W,
Strobel O et al. Small (Sendai negative) branch-duct IPMNs:
not harmless. Ann Surg 2012; 256: 313–320.
19 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.
20 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.
21 Dindo D, Demartines N, Clavien P-A. Classification of
surgical complications. Ann Surg 2004; 240: 205–213.
22 Strobel O, Büchler MW. Drainage after
pancreaticoduodenectomy: controversy revitalized. Ann Surg
2014; 259: 613–615.
23 Tol JA, Busch OR, van Delden OM, van Lienden KP, van
Gulik TM, Gouma DJ. Shifting role of operative and
nonoperative interventions in managing complications after
pancreatoduodenectomy: what is the preferred intervention?
Surgery 2014; 156: 622–631.
24 Allen PJ, Gönen M, Brennan MF, Bucknor AA,
Robinson LM, Pappas MM et al. Pasireotide for
postoperative pancreatic fistula. N Engl J Med 2014; 370:
2014–2022.
25 Sauvanet A, Gaujoux S, Blanc B, Couvelard A, Dokmak S,
Vullierme MP et al. Parenchyma-sparing pancreatectomy for
presumed noninvasive intraductal papillary mucinous
neoplasms of the pancreas. Ann Surg 2014; 260: 364–371.
www.bjs.co.uk BJS 2015; 102: 1258–1266