Pancreaticoduodenal Resection for Malignancy in a Low-volumeCenter: Long-term Outcomes from a Developing Country
Abu Bakar Hafeez Bhatti • Mohammad Aasim Yusuf •
Syed Ather Saeed Kazmi • Aamir Ali Syed
� Societe Internationale de Chirurgie 2014
Abstract
Background The technical complexity of pancreatic
resection has made it a specialized procedure performed in
high-volume centers. It has been shown that patients
operated on in high-volume pancreatobiliary centers have
fewer complications and better survival. The purpose of
this study was to share our experience with and report long-
term outcomes of pancreaticoduodenal resections per-
formed in a low-volume center in Pakistan.
Methods Data of patients who underwent pancreati-
coduodenal resection for adenocarcinoma at our institute
from 1999 to 2012 were reviewed. A total of 39 patients
were included in the study. Variables included patients’
clinical and histopathological characteristics. Outcome was
determined based on complication rate, 30- and 90-day
mortality, disease-free survival, and overall survival. For
survival analysis, Kaplan–Meier curves were used and
significance was determined using a log rank test. Uni-
variate Cox analysis was performed to determine signifi-
cant factors for multivariate analysis.
Results The majority of tumors [20 (51 %)] were mod-
erate grade, T1/T2 [20 (51 %)], ampullary adenocarcino-
mas [18 (46 %)]. Mean hospital stay was 14 ± 8 days. The
mean number of nodes removed was 13.9 ± 6.9, while
mean number of positive nodes was 1 ± 1.7. Expected
5-year overall survival and relapse-free survival were 38
and 48 %, respectively. Overall 5-year survival was sig-
nificantly different with respect to nodal involvement, i.e.,
47 vs. 28 % (P = 0.018). On univariate analysis, nodal
involvement was the only factor associated with an
increased risk of death (P = 0.02, hazard ratio [HR] 2.9,
confidence interval [CI] 1.1–7.8).
Conclusion Low-volume centers are an acceptable alter-
nate to high-volume centers for performing pancreati-
coduodenal resection in carefully selected patients. Efforts
should be directed at developing specialized hepatobiliary
centers in developing countries.
Introduction
Pancreaticoduodenal (PD) resection is one of the most
complex procedures in gastrointestinal surgery [1].
Although Whipple is credited with performing the first
successful pancreatic head resection for ampullary cancer,
it was Alexander Brunschwig who performed the first
successful pancreaticoduodenectomy for pancreatic cancer
[2, 3]. High morbidity and mortality rates plagued this
procedure throughout the 20th century, leading to questions
regarding its effectiveness [4, 5]. Developments in various
aspects of patient care and refinement in the surgical
technique have helped improve the outcomes of this
complex procedure and impressive results have been
achieved in the last decade. More recently, a mortality rate
as low as 1 % has been reported [1]. The morbidity rate is
still high and ranges between 30 and 66 % [1, 6–8]. An
inverse relationship between hospital volume and morbid-
ity and mortality has been observed [9]. High-volume
A. B. H. Bhatti (&) � A. A. Syed
Department of Surgical Oncology, Shaukat Khanum Memorial
Cancer Hospital and Research Centre, Lahore, Pakistan
e-mail: [email protected]
M. A. Yusuf
Department of Gastroenterology, Shaukat Khanum Memorial
Cancer Hospital and Research Centre, Lahore, Pakistan
S. A. S. Kazmi
Department of Medical Oncology, Shaukat Khanum Memorial
Cancer Hospital and Research Centre, Lahore, Pakistan
123
World J Surg
DOI 10.1007/s00268-014-2644-6
centers have reported superior results with pancreatico-
duodenectomy, not only with respect to postoperative
complications and mortality but also for long-term survival
[10–12]. In recent years, low-volume centers have also
demonstrated acceptable short- and long-term survival
rates for PD resections [13, 14]. Long-term survival data
from low-volume centers in developing countries is lim-
ited. Although no standard definition exists as to what
constitutes a low-volume center; hospitals that perform
fewer than five resections a year are generally considered
low volume. The purpose of this study was to share our
experience with and report long-term outcomes of PD
resection for malignancy performed at a low-volume center
in Pakistan.
Methods
Patient selection
A retrospective review of patients who underwent pan-
creatic resection for suspected malignancy from December
1999 to December 2012 was performed. A total of 39
patients who underwent PD resection and were diagnosed
with adenocarcinoma on final histopathology were inclu-
ded in the study. Three patients with histopathological
diagnosis of neuroendocrine tumor (1), intraductal papil-
lary mucinous neoplasm (1), and chronic pancreatitis (1)
were excluded to control for potential favorable effects on
survival.
Institutional details
Shaukat Khanum Cancer Hospital performs close to 50
upper and lower gastrointestinal operations each per year.
The number of PD procedures is low (average of \3 per
year) due to the advanced stage at presentation in a sig-
nificant number of patients which renders them inoperable.
All pancreatic surgeries were performed by a single sur-
geon and the most common technique employed was the
standard Whipple’s procedure. Figure 1 shows the man-
agement plan for patients treated at our institute. Patients
were initially seen at the walk-in clinic to determine eli-
gibility for treatment at SKM. If eligible, they were refer-
red to the Gastroenterology Service where essential lab,
diagnostic (if not already established), and staging workup
was performed. All patients underwent a CT scan with
pancreatic protocol for local and distant staging, and an
endoscopic ultrasound exam (starting in 2005) to determine
local resectability, particularly with respect to vascular
involvement. All patients with biochemical evidence of
obstructive jaundice underwent endoscopic retrograde
cholangiopancreatography (ERCP) and stent placement.
We tend to stent all our patients because many present with
advanced disease with more pronounced jaundice than is
usually the case in the West, and because we often have a
significant waiting time between diagnosis/workup and
surgery. It is our institutional policy to confirm a histologic
diagnosis in all our patients prior to initiation of cancer
treatment. Moreover, we have seen a few patients with
chronic pancreatitis and also with tuberculosis where it was
impossible to distinguish between cancer and chronic
inflammation on imaging. Therefore, we feel it appropriate
in our clinical environment to establish a histologic diag-
nosis with certainty prior to major surgery. All patients
were discussed in tumor boards involving surgeons, med-
ical and radiation oncologists, pathologists, gastroenterol-
ogists, and radiologists. Percutaneous transhepatic
cholangiographic drainage of the biliary system was per-
formed in patients in whom ERCP was unsuccessful. If
patients were deemed resectable, they were scheduled for
surgical resection. Criteria for resectability included no
distant metastases, no radiographic evidence of superior
mesenteric vein (SMV) or portal vein (PV) distortion, and
clear fat planes around the celiac axis, hepatic artery, and
SMA. Patients with borderline resectable/locally advanced
unresectable tumors were not considered for primary sur-
gery and were referred for downstaging with chemother-
apy. Tumors were considered borderline if there were no
distant metastases but distortion or narrowing of the SMV
or PV was present. Similarly, patients with hepatic artery
or superior mesenteric artery were not considered for pri-
mary surgery and referred for chemotherapy. Patients with
Initial workup
CT/EUS/ERCP
Tumor board
Resectable
Surgery +/-chemotherapy
Borderline resectable/locally
advanced
Preoperative chemotherapy followed by assessment
Resectable Irresectable
Metastatic
Palliative chemotherapy+/-palliative surgery
Fig. 1 Diagrammatic representation of management plan for patients
with periampullary tumors
World J Surg
123
metastatic disease and poor functional status were not
considered surgical candidates. Postoperatively, patients
were kept in the intensive care unit for 1 day and shifted to
the floor if stable. Patients were seen every 3 months in the
first postoperative year with a CT scan and tumor markers
performed at 6 months. Patients were seen 6 months into
the second postoperative year and yearly after that, with
tumor markers and CT scan performed every year. The
protocolized follow-up was conducted for early detection
of a local recurrence (which might be amenable to surgical
excision) and to assess general well-being of our patients.
Preoperative chemotherapy was used in patients with bor-
derline resectable/locally advanced disease; gemcitabine/
fluorouracil-based chemotherapy was commonly used.
Adjuvant chemotherapy was used in pathologically node-
positive tumors, poorly differentiated tumors, or tumors
with perineural invasion. The standard adjuvant chemo-
therapy was six cycles of gemcitabine. Radiotherapy was
used in a very few patients based on the recommendations
of the multidisciplinary meeting.
Statistical analysis
Patient characteristics, including demographics, clinico-
pathological details, treatment modalities, and postopera-
tive course were assessed. Patients were followed until
June 2013. The primary objective of the study was to
determine 5-year disease-free survival (DFS) and overall
survival (OS) of PD resection and to identify statistically
significant variables. In addition, we also evaluated post-
operative complications, their management, and 30- and
90-day mortality rates. DFS was calculated by subtracting
date of surgery from date of relapse. OS was calculated by
subtracting date of surgery from date of death or last fol-
low-up. Expected 5-year survival was calculated using
Kaplan–Meier survival curves, and the log rank test was
used to determine significance. For categorical variables,
Fischer’s exact test and the v2 test were used, while for
interval variables, the t-test was applied. Univariate Cox
analysis was performed to identify variables for multivar-
iate analysis. Our institutional ethics committee granted
exemption from formal IRB review of this study prior to
submission.
Results
Patient characteristics
The mean number of resections performed per year was
less than 3. Figure 2 shows the number of procedures
performed in each year of the study period. The median age
of the patients was 55 (30–72) years and the median
follow-up was 15 (0–140) months. The male:female ratio
was 2.5:1. The most common predominant symptom at
presentation was jaundice in 27 (69 %) patients. Majority
of patients had good functional status and were categorized
as American Society of Anesthesiologists (ASA) class I–II.
Ampullary adenocarcinoma was the most common under-
lying malignancy, in 18 (46 %) patients. Preoperative
diagnosis of adenocarcinoma was established in 37 (95 %)
patients. Table 1 presents the demographics, clinical
details, and diagnostic modalities used in patients.
Treatment modalities
The standard Whipple procedure, used in 35 (90 %)
patients, was the most common pancreatic resection pro-
cedure. Two patients with colonic adenocarcinoma under-
went unplanned PD resection along with colonic resection.
One patient underwent a Whipple procedure while the
other required total pancreatectomy and right nephrectomy.
Table 2 presents the various treatment options used. Che-
motherapy was used in 14 (36 %) patients. In 11 (28 %)
patients, it was used as adjuvant treatment, while 2 (5 %)
patients received preoperative chemotherapy. One patient
received both neoadjuvant and adjuvant chemotherapy.
This was a 35-year-old patient with a locally advanced
tumor at presentation in whom adjuvant chemotherapy was
given to maximize chances of cure after successful surgery.
Preoperative combined chemoradiotherapy was used in one
patient with initial vascular involvement but good func-
tional status. In this patient, 5-fluorouracil-based chemo-
therapy and 50-Gy radiation in 25 fractions were given
Fig. 2 Number of pancreaticoduodenectomies in each year of the
study period
World J Surg
123
before surgical resection. Adjuvant radiation was given in a
patient with margin-positive resection and in another with
involvement of bile duct serosa with tumor cells. Palliative
radiotherapy was used in one patient with painful sacroiliac
metastasis in the first year of follow-up after surgery.
Histopathological variables
The majority of patients [31/36 (86 %)] had well to
moderately differentiated tumors. Lymphovascular [3/29
(10 %)] or perineural invasion [9/29 (31 %)] was identi-
fied in a small number of patients. Tumors were equally
distributed in terms of nodal involvement [N0 vs. N1 (54
vs. 46 %)] and tumor size [T1/T2 vs. T3/T4 (51 vs.
49 %)]. Margins were negative in all but one patient. For
the actual number of observed relapses, the distribution of
histopathological variables was not significantly different.
For actual number of observed mortalities, nodal
involvement was the only significant variable (P = 0.008)
(Table 3).
Complications and 30-day and 90-day mortality rates
Seventeen patients developed 19 (48.7 %) postoperative
complications. There were three deaths within 30 days for
a mortality rate of 7.6 % and six deaths within 90 days for
a rate of 15 %. The overall complication rate was 48.7 %.
Common complications included pancreaticojejunal (PJ)
leak in seven patients and wound infection in four patients.
Table 4 presents the various complications, their manage-
ment, and success rates. Three patients required completion
pancreatectomy for management of a PJ leak due to
Table 1 Patient characteristics and diagnostic modalities in study
group
n %
Age (years)
(mean ± SD)
54.9 ± 10.8
Gender Male 28 71.1
Female 11 28.9
ASA classa I 5 14.7
II 19 55.9
III 10 29.4
Predominant
symptom
Jaundice 27 69.2
Abdominal pain 9 23
Others 3 7.5
Preoperative ERCP Yes 34 87.1
No 5 12.9
Endoscopic
ultrasound
Yes 27 69.2
No 12 30.8
Preoperative
diagnosisaERCP ? biopsy 27 73
ERCP ? brush cytology 1 2.7
OGD ? biopsy 3 8.1
EUS ? biopsy 3 8.1
Percutaneous abdominal US-
guided biopsy
1 2.7
Colonoscopy ? biopsy 2 5.4
Diagnosis Ampullary adenocarcinoma 18 46.2
Duodenal adenocarcinoma 12 30.8
Pancreatic adenocarcinoma 7 17.9
Colonic adenocarcinoma 2 5.1
a Variables with missing data
Table 2 Various treatment modalities used in the study group
n %
Surgical
procedure
Whipple’s procedure 35 89.7
Total pancreatectomy 2 5.1
Right hemicolectomy ? Whipple’s 1 2.6
Right hemicolectomy ? right
nephrectomy ? total
pancreatosplenectomy
1 2.6
Chemotherapy Adjuvant 12 75
Neoadjuvant 2 12.5
Palliative 2 12.5
Radiation
therapy
Adjuvant 3 60
Neoadjuvant 1 20
Palliative 1 20
Table 3 Histopathological variables and their significance with
respect to disease relapse and death
n P value
Disease
relapse
Mortality
Gradea Well 11 NS NS
Moderate 20
Poor 5
p Tumor T1/T2 20 NS NS
T3/T4 19
p Nodal
involvement
N0 21 NS 0.008
N1 18
Nodes removed Mean 13.9 ? 6.9 NS NS
Nodes positive Mean 1 ? 1.7 NS NS
Margins Negative 38 NS NS
Positive 1
LV invasiona Present 3 NS NS
Absent 26
Perineural
invasionaPresent 9 NS NS
Absent 20
P value represented in italic is not statistically significant (P C 0.05 )
NS not significant, p pathologicala Variables with missing data
World J Surg
123
complete disruption of anastomosis. There were two
patients with partial disruption of the PJ anastomosis that
required resuturing of the disruption point in one patient
and omental buttress repair in the other. Superior mesen-
teric artery repair was performed in one patient with early
postoperative intra-abdominal bleeding. There was one
patient with a choledochojejunostomy leak that was man-
aged successfully with T-tube repair. One patient devel-
oped a myocardial infarct in the postoperative period that
was managed successfully with medical treatment.
Disease-free and overall survival
Fifteen (38 %) patients had a disease relapse, with liver
being the first site of relapse in 11 (73 %) patients. Other
common relapse sites were lungs and para-aortic lymph
nodes in two patients each (Table 5). Median recurrence-
free survival was 9 (0–140) months. The expected 5-year
DFS was 48 %. There were 22 deaths in the follow-up
period. The median OS was 12 (0–140) months. The
expected 5-year OS was 38 %. Overall 5-year survival was
significantly different with respect to nodal involvement
(N0 vs. N1), i.e., 47 vs. 28 % (P = 0.018), as shown in
Fig. 3. Median OS for pancreatic versus nonpancreatic
adenocarcinoma was 10 (1–108) and 16 (0–140) months,
respectively. On univariate analysis, nodal involvement
was the only factor associated with a decrease in OS.
Patients with nodal involvement had threefold increase in
the risk of death (P = 0.02, hazard ratio [HR] 2.9, confi-
dence interval [CI] 1.1–7.8). Multivariate analysis could
not be performed as nodal involvement was the only sig-
nificant variable on univariate analysis.
Discussion
A number of studies have reported outcomes of PD
resections in low-volume centers [13–16]. They have
shown acceptable outcome for this complex procedure
when compared with specialized centers. Similar results
have been reported in the current study, with acceptable
Table 4 Postoperative complications with subsequent management
and success rates
n Management n Success
rate
(%)
Pancreaticojejunal leak 7 Completion
pancreatectomy
3 57
Omental buttress
repair
1
Conservative 1
Refashioning 1
CT-guided aspiration 1
Choledochojejunostomy
leak
1 T-tube repair 1 100
Wound infection 4 Antibiotics 3 100
100Debridement 1
Gastroparesis 2 Conservative 1 100
Stomal ulceration 1 Injection
sclerotherapy
1 100
Myocardial infarct 1 Medical management 1 100
Intra-abdominal
hematoma
1 Superior mesenteric
artery repair
1 0
Intra-abdominal
collection
2 CT-guided drainage 1 100
Conservative 1
Overall complications 19 19 79
Table 5 Pattern of relapse in patients after pancreaticoduodenal
resection for adenocarcinoma
Site of recurrence n %
Liver 8 15
Lungs 2 5
Para-aortic lymph nodes 1 2.5
Bones 1 2.5
Liver ? para-aortic nodes 1 2.5
Liver ? surgical bed 1 2.5
Liver ? lungs 1 2.5
Overall 15 38.4
Nodal involveNo nodinvolve
Log ran
ement dal ement
nk P value:
1 year 2
41
78
: 0.01
2 year 3 y
24
70
year 4 y
24
70
year 5
24
70
year
24
47
Fig. 3 Kaplan–Meier survival curves for overall survival with
respect to nodal involvement in patients who underwent pancreati-
coduodenectomy for adenocarcinoma
World J Surg
123
short-term outcomes and long-term survival. The unique
features of the present study were the high percentage of
patients with ampullary/duodenal carcinoma and exclusion
of patients with benign/borderline tumors or patients who
did not have pancreatic head resection. Limitations of the
study included the retrospective nature of study, small
sample size, and relatively short follow-up.
The overall complication rate of \50 %, mortality rate
close to 7 %, and mean hospital stay of 14 days parallels
the results from many low- and some high-volume centers
[6, 9, 13–19]. In the last decade, low-volume centers have
reported a morbidity rate close to 60 % and a mortality rate
around 10 % [6, 14, 20]. Despite improvements in short-
term outcomes of pancreatic resections, long-term results
after surgical resection for pancreatic cancer remain
unchanged, with a 5-year survival rate of 20–25 % [14, 21–
23]. Survival for ampullary cancer is better at 40–60 % for
low- and high-volume centers [15, 24, 25]. Significant
factors reported to affect outcome after pancreatic resection
for malignancy include tumor stage, margin status, grade,
nodal involvement, perineural invasion, and underlying
pathology [26–28]. A high number of nonpancreatic can-
cers probably led to the better OS rate in the current study.
There is no consensus on how margin status is defined in
pancreatic cancer. In North America, tumor cells at the
surface of resection margins are considered positive, while
at other places tumor cells within 1 mm of the resection
margin are taken as positive [29–31]. It has been shown
that a negative margin of [1.5 mm translates into
improved survival outcomes in patients with pancreatic
resection for malignancy [32]. In the present study, no
tumor cells on resected margins microscopically were
considered negative and nearly all patients had negative
margins. Since pancreatic malignancy was not the most
common underlying malignancy, the effect of negative
margins on survival becomes less important. Nodal
involvement was identified as the only independent factor
significantly affecting survival. Perineural and lympho-
vascular invasion was identified in a small number of
patients and its significance could not be determined.
Indications for adjuvant chemotherapy include margin-
positive disease or nodal positivity. For locally advanced
tumors, chemoradiotherapy in a neoadjuvant setting is
being used extensively [33]. A recent meta-analysis
reported improved 12-month survival in patients who
received postoperative gemcitabine in combination with
radiotherapy [34]. Nearly 40 % patients in the present
study received chemo- or radiotherapy.
Schell et al. [14] compared outcomes following PD
resections between high- and low-volume centers affiliated
with the University of California. Over a period of 15 years
the postoperative complication rate (58.8 vs. 63.1 %),
mortality rate (4 vs. 4 %) and 5-year survival rate (19 vs.
18.3 %) were not significantly different between the two
groups. Although no definition of negative margins was
provided, positive margins were encountered in 25 % of
surgical specimens in both high- and low-volume centers.
Patients with a broad range of underlying pathologies like
adenocarcinoma, neuroendocrine tumors, intraductal pap-
illary mucinous neoplasm (IPMN), and chronic pancreatitis
were included. They concluded that low-volume centers
can achieve comparable results provided the expertise and
care facilities match those of high-volume centers. The
authors believed that the matching was possible due to
continuous sharing of operative techniques and periopera-
tive care pathways. Variable perioperative mortality rates
have been reported from low-volume centers, reaching up
to 20 % [6, 14, 18, 20]. A systematic review reported a
perioperative mortality rate of 5.5 vs. 11 % in high- and
low-volume centers, respectively [20]. Another systematic
review reported the postoperative mortality rate in hospi-
tals that performed fewer than five resections to be as high
as 16 % [35]. A complication rate of \50 % and 30-day
mortality rate of 7.6 % in the present study is comparable
to those of many low-volume centers [6, 14, 20].
In the current study, a 5-year DFS rate of 48 % and OS
rate of 38 % were observed. The high number of periam-
pullary tumors probably accounts for the improved survival
outcome. Patients with pathologies other than adenocarci-
noma were excluded from the study and thus issues of
selection bias were addressed and survival was determined
specifically for periampullary adenocarcinoma. Surgical
margins were positive in only one patient. We feel that this
low rate of margin positivity was due to surgical standards.
However, the variable definitions of margin positivity in the
literature and the high number of pancreatic head cancers,
which increase the risk of positive pancreatic resection
margins, in other studies may have confounded the results
of these studies. Uniformity in surgical technique (all pro-
cedures were performed by a single surgeon), careful
patient selection ([60 % of patients were ASA I and II),
absence of complex vascular resections, and accurate pre-
operative planning ([94 % of patients had a preoperative
tissue diagnosis, [85 % had preoperative stenting, and
70 % had preoperative EUS staging) could have contributed
to the improved short- and long-term outcomes that we
report here. Not all low-volume centers are as well equipped
as ours, follow a uniform surgical technique, and have
stringent patient selection criteria. Metreveli et al. [6]
reported outcomes from a low-volume community hospital.
Twenty-five surgeons performed a total of 63 pancreatic
resections; EUS was performed in only 11 % and ERCP in
44 % of patients. Preoperative tissue diagnosis was estab-
lished in 41 % of patients included in their study. Although
acceptable results were achieved in the current study, the
patient cohort was very selective and several issues remain
World J Surg
123
unanswered. Low-volume centers may provide acceptable
outcomes in carefully selected patients who have good
functional status and localized tumors not requiring com-
plex resections, but how are the high-risk surgical candi-
dates with vascular/partial arterial involvement managed?
In Pakistan, there was no center performing complex vas-
cular resections during the study period. Since most patients
investigated at Shaukat Khanum Hospital could not afford
treatment, travel outside the country was not possible for
these patients. We treated these patients with chemotherapy
with the intention of downstaging them in the hope that their
tumors became resectable. Unfortunately, there were very
few patients who ultimately had a pancreaticoduodenecto-
my. For the majority of low-volume centers, there is sub-
stantial room for improvement. The limitations in
preoperative planning, early detection of complications, and
standardization of surgical technique in low-volume centers
cannot be overemphasized. The present study showed that if
these limitations can be addressed, results comparable to
those of high-volume centers can be achieved for a specific
patient group.
Conclusion
The present study reports survival outcomes in addition to
short-term morbidity and mortality rates in patients who
underwent PD resection for malignancy in a low-volume
center for pancreatobiliary procedures. To our knowledge,
this is the first study reporting long-term survival in a
Pakistani cohort after PD resection for adenocarcinoma.
The study demonstrates that low-volume centers can
achieve acceptable results for complex PD resection in a
carefully selected patient cohort. The need to develop
specialized centers with high volume cannot be overem-
phasized since surgical resection is the only form of cure
for this aggressive malignancy.
Conflicts of interest The authors have no conflicts of interest or
financial ties to disclose.
References
1. Winter JM, Cameron JL, Campbell KA et al (2006) 1423 pan-
creaticoduodenectomies for pancreatic cancer: a single-institution
experience. J Gastrointest Surg 10:1199–1210 discussion
1210–1211
2. Whipple AO, Parson N, Mullins C (1935) Treatment of carci-
noma of the ampulla of vater. Ann Surg 102:763–779
3. Brunschwig A (1937) Resection of the head of the pancreas and
duodenum for carcinoma. Pancreatoduodenectomy. Surg Gyne-
col Obstet 65:681–684
4. Crile J Jr (1970) The advantages of bypass operations over radical
pancreatoduodenectomy in the treatment of pancreatic carci-
noma. Surg Gynecol Obstet 130:1049–1053
5. Gudjonsson B (1987) Cancer of the pancreas. 50 years of sur-
gery. Cancer 60:2284–2303
6. Metreveli RE, Sahm K, Abdel-Misih R et al (2007) Major pan-
creatic resections for suspected cancer in a community-based
teaching hospital: lessons learned. J Surg Oncol 95:201–206
7. Traverso LW, Shinchi H, Low DE (2004) Useful benchmarks to
evaluate outcomes after esophagectomy and pancreaticoduoden-
ectomy. Am J Surg 187:604–608
8. Bassi C, Dervenis C, Butturini G et al (2005) Post operative
pancreatic fistula: an international study group (1 SGPF) defini-
tion. Surgery 138:8–13
9. Sosa JA, Bowman HM, Gordon TA et al (1998) Importance of
hospital volume in the overall management of pancreatic cancer.
Ann Surg 228:429–438
10. Van Oost FJ, Luiten EJ, van de Poll-Franse LV et al (2006)
Outcome of surgical treatment of pancreatic, peri-ampullary and
ampullary cancer diagnosed in the south of The Netherlands: a
cancer registry based study. Eur J Surg Oncol 32:548–552
11. Fong Y, Gonen M, Rubin D et al (2005) Long-term survival is
superior after resection for cancer in high-volume centers. Ann
Surg 242:540–544 discussion 544–547
12. Parks RW, Bettschart V, Frame S et al (2004) Benefits of spe-
cialisation in the management of pancreatic cancer: results of a
Scottish population-based study. Br J Cancer 91:459–465
13. Afsari A, Zhandoug Z, Young S et al (2002) Outcome analysis of
pancreaticoduodenectomy at a community hospital. Am Surg
68:281–284 discussion 284–285
14. Schell MT, Barcia A, Spitzer AL et al (2008) Pancreaticoduo-
denectomy: volume is not associated with outcome within an
academic health care system. HPB Surg 2008:825940
15. Choi SB, Kim WB, Song TJ et al (2011) Surgical outcomes and
prognostic factors for ampulla of Vater cancer. Scand J Surg
100:92–98
16. Mukhtar RA, Kattan OM, Harris HW (2008) Variation in annual
volume at a university hospital does not predict mortality for
pancreatic resections. HPB Surg 2008:190914
17. Finlayson EV, Birkmeyer JD (2003) Effects of hospital volume
on life expectancy after selected cancer operations in older adults:
a decision analysis. J Am Coll Surg 196:410–417
18. Nordback L, Parviainen M, Raty S et al (2002) Resection of the
head of the pancreas in Finland: effects of hospital and surgeon
on short-term and long-term results. Scand J Gastroenterol
37:1454–1460
19. Garcea G, Dennison AR, Pattenden CJ et al (2008) Survival
following curative resection for pancreatic ductal adenocarci-
noma. A systematic review of the literature. JOP 9:99–132
20. Pal KM, Bari H, Nasim S (2011) Pancreaticoduodenectomy: a
developing country perspective. J Pak Med Assoc 61:232–235
21. Cooperman AM (2001) Pancreatic cancer: the bigger picture.
Surg Clin North Am 81:557–574
22. Richter A, Niedergethmann M, Sturm JW et al (2003) Long-term
results of partial pancreaticoduodenectomy for ductal adenocar-
cinoma of the pancreatic head: 25-year experience. World J Surg
27:324–329
23. Yeo CJ, Cameron JL, Sohn TA et al (1997) Six hundred fifty
consecutive pancreaticoduodenectomies in the 1990s: pathology,
complications, and outcomes. Ann Surg 226:248–257 discussion
257–260
24. Sudo T, Murakami Y, Uemura K et al (2008) Prognostic impact
of perineural invasion following pancreatoduodenectomy with
lymphadenectomy for ampullary carcinoma. Dig Dis Sci
53:2281–2286
World J Surg
123
25. Qiao QL, Zhao YG, Ye ML et al (2007) Carcinoma of the
ampulla of Vater: factors influencing long-term survival of 127
patients with resection. World J Surg 31:137–143 discussion
144–146
26. Sohn TA, Yeo CJ, Cameron JL et al (2000) Resected adenocar-
cinoma of the pancreas—616 patients: results, outcomes, and
prognostic indicators. J Gastrointest Surg 4:567–579
27. Van Roest MH, Gouw AS, Peeters PM et al (2008) Results of
pancreaticoduodenectomy in patients with periampullary adeno-
carcinoma: perineural growth more important prognostic factor
than tumor localization. Ann Surg 248:97–103
28. Wagner M, Redaelli C, Lietz M et al (2004) Curative resection is
the single most important factor determining outcome in patients
with pancreatic adenocarcinoma. Br J Surg 91:586–594
29. Raut CP, Tseng JF, Sun CC et al (2007) Impact of resection status
on pattern of failure and survival after pancreaticoduodenectomy
for pancreatic adenocarcinoma. Ann Surg 246:52–60
30. Esposito I, Kleeff J, Bergmann F et al (2008) Most pancreatic
cancer resections are R1 resections. Surg Oncol 15:1651–1660
31. Verbeke CS (2008) Resection margins and R1 rates in pancreatic
cancer–are we there yet? Histopathology 52:787–796
32. Chang DK, Johns AL, Merrett ND et al (2009) Margin clearance
and outcome in resected pancreatic cancer. J Clin Oncol
27:2855–2862
33. Shrikhande SV, Barreto SG (2012) Surgery for pancreatic car-
cinoma: state of the art. Indian J Surg 74:79–86
34. Zhu CP, Shi J, Chen YX et al (2011) Gemcitabine in the che-
moradiotherapy for locally advanced pancreatic cancer: a meta-
analysis. Radiother Oncol 99:108–113
35. Van Heek NT, Kuhlmann KF, Scholten RJ et al (2005) Hospital
volume and mortality after pancreatic resection: a systematic
review and an evaluation of intervention in the Netherlands. Ann
Surg 6:781–788 discussion 788–790
World J Surg
123