Surgical treatment of pancreatic adenocarcinoma: actual survival
and prognostic factors in 343 patients
K.F.D. Kuhlmanna, S.M.M. de Castroa, J.G. Wesselingb, F.J.W. ten Katec,
G.J.A. Offerhausc, O.R.C. Buscha, T.M. van Gulika, H. Obertopa, D.J. Goumaa,*
aDepartment of Surgery, Academic Medical Center from the University of Amsterdam, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands
cDepartment of Pathology, Academic Medical Center from the University of Amsterdam, Amsterdam, The Netherlands
Received 30 September 2003; accepted 3 October 2003
Survival data of patients with pancreatic carcinoma are often overestimated because of incomplete follow-up. Therefore, the aim
of this study was to approach complete follow-up and to analyse survival and prognostic factors of patients who underwent surgical
treatment for pancreatic adenocarcinoma. Between 1992 and 2002, 343 patients underwent surgical treatment for pancreatic
adenocarcinoma. One hundred and sixty patients underwent a resection with a curative intention and 183 patients underwent
bypass surgery for palliation. Follow-up was complete for 93% of patients. Median survival after resection and bypass was 17.0 and
7.5 months, and 5-year survival was 8% and 0, respectively. In multivariate analysis, tumour-positive lymph nodes, non-radical
surgery, poor tumour differentiation, and tumour size were independent prognostic factors for survival after resection. For patients
treated with bypass surgery, metastatic disease and tumour size independently predicted survival. In conclusion, actual survival of
patients with pancreatic adenocarcinoma is disappointing compared with the actuarial survival rates reported in the literature. The
independent prognostic factors for survival of patients who underwent surgical treatment for pancreatic adenocarcinoma are
# 2003 Elsevier Ltd. All rights reserved.
Keywords: Pancreatic ductal adenocarcinoma; Adenocarcinoma; Surgery; Pancreaticoduodenectomy; Gastric bypass; Biliary bypass; Palliative care;
Survival; Prognostic factors
Pancreatic cancer is a highly lethal disease. The
incidence in the United States of America (USA) and
Western Europe is 10/100000 per year and approaches
mortality. The overall 5-year survival is less than 4%
and has hardly improved over the last two decades .
Pancreaticoduodenectomy is still the only treatment
with a curative potential. However, 80% of patients are
not eligible for surgical resection because of local spread
or metastatic disease at the time of diagnosis [2,3]. Thus,
most patients are palliatively treated to improve the
quality of their remaining life.
Endoscopic or transhepatic stenting are effective for
short-term relief of jaundice, but stent exchanges are
needed frequently and therefore cause considerable
morbidity . Surgical bypass procedures (biliary, gas-
tric or both) can offer optimal long-term palliation of
obstructive jaundice and can prevent or treat gastro-
intestinal obstruction . Furthermore, a chemical
coeliac plexus nerve block can easily be performed during
laparotomy and can effectively manage pain in patients
with unresectable pancreatic cancer .
Survival rates and prognostic factors for patients who
underwent surgical treatment for pancreatic adeno-
carcinoma have been reported before. However, many
studies include a limited number of patients, cover a
long period of time, analyse all periampullary tumours
(including ampullary, distal bile duct and duodenal
tumours) or have limited follow-up with an estimated
0959-8049/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
European Journal of Cancer 40 (2004) 549–558
* Corresponding author. Tel.: +31-20-5662166; fax: +31-20-
E-mail address: firstname.lastname@example.org (D.J. Gouma).
survival [7–19]. These estimated actuarial curves may
show better survival results than the actual survival.
The present study represents a large single centre ser-
ies including 343 patients who underwent surgical
treatment for histologically-proven pancreatic adeno-
carcinoma. Only 23 patients were alive at the end of
follow-up and therefore Kaplan–Meier survival curves
approach the actual survival. Earlier studies of our
department have focused on the survival and the prog-
nostic factors of periampullary tumours, but not on
pancreatic adenocarcinoma alone [13,15,20–22]. There-
fore, our aim was to evaluate the survival of patients
with histologically-proven pancreatic adenocarcinoma
and to analyse the prognostic factors for survival after
resection or bypass surgery at the time of diagnosis.
2. Patients and methods
In the 10-year period from January 1992 to December
2001, a total of 759 consecutive patients underwent
surgical treatment (resection in 459 patients and bypass
in 300 patients) for a periampullary mass. In 358 of 759
patients pancreatic adenocarcinoma was confirmed
histologically. Fifteen patients, who were initially
treated elsewhere, were referred to our centre in a late
stage of the disease for symptoms of gastric obstruction.
These 15 patients were excluded. The remaining 343
patients were selected and data were obtained from a
prospectively collected database.
2.1. Operative procedures
Of these 343 patients, 329 (96%) underwent an
exploratory laparotomy with the intention to perform a
resection. The indications to perform a bypass were
metastatic disease or locally advanced tumour due to
involvement of the portal vein, major arteries or infil-
tration of the mesocolon or retroperitoneum. When
irresectability was assessed at exploratory laparotomy,
histological conformation was obtained using frozen
sections. The remaining 14 patients directly underwent
bypass surgery because the tumour proved irresectable
at diagnostic work-up. In these patients, non-surgical
palliation was not considered because symptoms of
gastric obstruction were present.
Of the 343 patients, 160 patients underwent a stan-
dard resection (group 1) as described in Ref. . A
pylorus-preserving pancreaticoduodenectomy was per-
formed in 133 and a classic Whipple’s procedure in 27
patients. Reconstruction was performed with a single
retrocolic jejunal limb reconstruction. In 33 patients
with minimal venous ingrowth, portal or superior
mesenteric vein resection was performed.
Hundred and eighty three patients underwent pallia-
tive bypass surgery (group 2), 157 a double bypass
(hepaticojejunostomy and gastroenterostomy) and 26
patients a single bypass (hepaticojejunostomy: n=15,
gastroenterostomy: n=10, and pancreaticojejunostomy:
n=1). Ninety patients underwent bypass surgery for
metastatic disease and 93 patients for a locally advanced
During surgery, frozen sections of the resection mar-
gins were taken for histological examination. A frozen
section was taken of the pancreatic resection margin in
22% of the patients and in 43% of the patients of the
bile duct margin. Frozen sections of the other resection
margins (superior mesenteric artery, portal vein and
duodenal resection margins) were only taken if indi-
cated. In the case of a tumour-positive frozen section,
re-resection was performed.
In the operating room, the surgeon marked the resec-
tion margins at the resection specimen at the level of bile
duct, pancreatic duct, superior mesenteric artery and
portal vein margin. Resection specimens were analysed
for location, size and differentiation of the tumour and
for the status of lymph nodes and resection margins. In
the bypass group, the tumour size was assessed by pre-
operative imaging or was estimated during the oper-
ation. The histological tumour differentiation in this
group was assessed in biopsies of the tumour.
Tumours were categorised according to the American
When margins at the superior mesenteric artery or
portal vein were positive, tumours were classified as
stage IVa, even when ingrowth could not be proven
2.3. Patient evaluation and follow-up
characteristics, symptoms at the time of operation and
pre-operative American Society of Anesthesiologists
(ASA) score. Peri-operative parameters included surgi-
cal complications: anastomotic leakage (more than three
times serum amylase or leakage proven at laparotomy),
haemorrhage, abdominal abscess, wound infection, and
delayed gastric emptying (stomach drainage for longer
than 10 days postoperatively or intolerance for normal
food intake for longer than 2 weeks postoperatively),
general complications (e.g., cardio-pulmonary compli-
cations and urinary tract infections), relaparotomy dur-
ing the admission for initial surgery, post-operative
hospital stay, in-hospital mortality, blood transfusion
(defined as packed red blood cells (PRC) within the first
24 h postoperatively and during hospital stay) and
adjuvant or palliative chemotherapy and/or radio-
therapy. All patients were followed for survival analysis
550 K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558
on a regular basis in the outpatient clinic. Follow-up
ended in April 2003. When follow-up was incomplete,
the general practitioner or referring physicians outside
our centre were contacted.
Differences between group 1 and 2 were analysed
applying Chi-square test, ANOVA or Kruskall–Wallis
test. Survival was measured from the day of surgery
until death or until the last day of the follow-up period.
Hospital mortality was included in the survival analysis.
Survival was analysed with the Kaplan–Meier method.
The log-rank test was used to compare differences in
survival between the groups. Statistical significance was
considered if P<0.05. After univariate analysis, all sig-
nificant prognostic factors were entered into a Cox
regression model to determine independent predictors of
survival. Statistical analysis was performed using Sta-
tistical Package for the Social Sciences (SPSS) version
10.1.0 (SPSS Inc., Chicago, IL).
3.1. Patient population
The two groups (group 1: resection and group 2:
bypass) were comparable in terms of age and sex dis-
tribution, pre-operative ASA-score and jaundice at the
time of presentation. In group 2, significantly more
patients presented with pain (P=0.01). Adjuvant or
palliative chemo- and radiotherapy were not equally
distributed over the two groups because of the different
regimes followed during the 10-year inclusion period.
Patients’ characteristics are summarised in Table 1.
3.2. Peri-operative parameters
Median post-operative hospital stay was significantly
longer in group 1 (20 versus 12 days with a range of 8–
222 versus 4–43, respectively). There was no hospital
mortality in group 1 and 2% (4 patients) in group 2.
One patient died of cardiac complications, one of pul-
monary complications due to carcinomatous pleuritis,
one of abdominal sepsis, and one of pulmonary insuffi-
ciency due to pneumonia.
Anastomotic leakage (7% versus 1%, P=0.01),
abdominal abscess (10% versus 3%, P=0.01) and gas-
trointestinal obstruction (14% versus 7%, P=0.03)
occurred significantly more frequently after resection.
Haemorrhage, wound infection, and general compli-
cations were not significantly different between groups.
One or more relaparotomies were performed in 9% of
patients in group 1 and in 4% of patients in group 2,
which was not significantly different (P=0.07). Patients
in group 1 received significantly more packed cells in the
first 24 h (1.1 versus 0.2 units of packed cells, P<0.01)
and during their hospital stay (2.2 versus 0.5 units
of packed cells, P<0.01) than patients in group 2.
Peri-operative parameters are summarised in Table 2.
Six of 35 frozen sections of the pancreatic resection
margin (17%) and two of 69 frozen sections of the bile
duct margin (3%) were tumour-positive and subsequent
re-resection was performed.
Resection group 1 (n=160)Bypass group 2 (n=183)P-value
Age (years) (median (range))
Pain (abdominal and/or back)
Adjuvant or palliative therapy
64 (43–84) 62 (33–78)NS
Numbers in parentheses represent percentages or range. ASA: American Society of Anesthesiologists, NS, non-significant (P value50.05).
aSome data are missing in these subgroups.
K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558551
The tumour diameter in group 1 was significantly
smaller than in group 2 (2.8 versus 3.4 cm, P<0.01).
There was no difference in tumour differentiation
between the two groups. In group 1, 68% of the patients
had one or more positive lymph node(s) in the resection
specimen. Fifty per cent of patients had tumour-positive
resection margins. The margin at the superior mesen-
teric artery was tumour-positive in 37 patients, in 36
patients the dissection planes, in 31 patients the pan-
creatic resection margin, in 29 patients the margin at the
portal vein, in 4 the resection margin of the bile duct
and in 3 patients the resection margin of the proximal
duodenum was tumour-positive. In group 2, metastatic
disease was present in 49%.
Resection group 1 (n=160) Bypass group 2 (n=183)P-value
Postoperative stay days (median)20 (8–222) 12 (4–43)
Hospital mortality0 (0) 4 (2)NS
Delayed gastric emptying
Relaparotomy15 (9) 8 (4) NS
PRC units (median)
First 24 h
During hospital stay
Numbers in parentheses represent percentages or range. PRC, packed red blood cell; NS, non-significant (P value50.05).
Pathology and tumour characteristics
Resection group 1 (n=160) Bypass group 2 (n=183)P-value
Tumour size (cm) (median (range))
Margins not well defined
Tumour-positive lymph nodes109 (68)ND–
Tumour-positive resection margins80 (50)––
Metastatic disease090 (49)
Numbers in parentheses represent percentages or range. NS, Non-significant (P value50.05); ND, not determined.
552 K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558
In group 1, most patients had stage III (49%) or stage
VIa (29%) disease according to the AJCC/UICC 1997
classification. In group 2, there was an equal distribu-
tion between stage VIa and VIb disease (Table 3).
3.4. Follow-up and survival
At the end of follow-up in April 2003, 22 patients
(6%) were alive (all in group 1) and 320 patients (93%)
were followed until death. One patient was lost to fol-
low-up in group 1 due to emigration. The median fol-
low-up was 16.9 months in group 1 and 7.5 months in
group 2. Median follow-up of the living patients was
31.8 months (17.3–134.0 months).
Median survival for group 1 was 17.0 months and
for group 2 7.5 months, with a 1, 2 and 5-year survi-
vals of 64%, 34% and 8% in group 1 and 27%, 3%
and 0 in group 2. Fig. 1 shows the Kaplan–Meier
survival curves for the two groups. In the survival curve
of group 1, the 23 patients that were alive at the end of
follow-up or lost to follow-up were censored as depic-
ted. There were seven actual 5-year survivors in this
series (range 60.2–134.0 months). The survival of all
patients divided by the AJCC/UICC 1997 classification
is shown in Fig. 2.
3.5. Prognostic factors for survival (univariate analysis)
In group 1, 4 of the 24 analysed variables (tumour
size, tumour differentiation, lymph nodes status and
resection margins) significantly influenced survival. In
group 2, 10 variables influenced survival: radiotherapy,
post-operative hospital stay, haemorrhage, abdominal
abscess, delayed gastric emptying, relaparotomy, more
than four PRC’s during admission, metastatic disease,
tumour size and UICC classification.
Fig. 1. Kaplan–Meier survival curves of patients after resection and
bypass. Censored patients are indicated with a cross. P-value: <0.01.
Fig. 2. Kaplan–Meier survival curves of all patients who underwent
surgery for pancreatic adenocarcinoma (resection and bypass) divided
by the AJCC/UICC 1997 classification. Overall P-value: <0.01.
Multivariate analysis of prognostic factors predicting favourable survival
Tested factor Reference factorHazard ratio (95% Confidence Interval)P-value
Group 1 (resection):
Tumour-positive resection margins
Poor tumour differentiation
1 cm decrease
Group 2 (bypass):
Metastatic disease present
Delayed gastric emptying present
Abdominal abscess present
1 cm decrease
Bold: independent prognostic factors.
K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558553
Fig. 3. Kaplan–Meier survival curves of patients who underwent a resection (group 1) for pancreatic adenocarcinoma according to the independent
prognostic factors. a: lymph nodes, b: resection margins, c: tumour differentiation (1=poor, 2=poor-intermediate, 3=intermediate, 4=inter-
mediate-high, 5=high) and d: tumour size in cm (patients with a tumour size >6 cm (n=2) are not depicted).
Fig. 4. Kaplan–Meier survival curves of patients who underwent a bypass (group 2) for pancreatic adenocarcinoma according to the independent
prognostic factors. a: metastatic disease, b: tumour size (patients with a tumour size <2 cm (n=5) and >6 cm (n=4) are not depicted).
554 K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558
3.6. Independent prognostic factors for survival
(Cox regression model)
For group 1, resection margins, lymph node status,
tumour differentiation and tumour size were entered in
the multivariate analysis. These four variables all inde-
pendently predicted the survival. One or more tumour-
positive lymph nodes predicted a poor survival with a
hazard ratio of 1.58 (95% Confidence Interval (CI)
1.08–2.31), tumour-positive resection margins with a
hazard ratio of 1.57 (CI 1.10–2.23), low tumour differ-
entiation with a hazard ratio of 1.31 (CI 1.13–1.52) and
tumour size (compared with a tumour that is 1 cm
smaller) with a hazard ratio of 1.25 (CI 1.06–1.48)
(Table 4 and Fig. 3).
For group 2, seven of 10 variables that were sig-
nificant in the univariate analysis were entered in the
multivariate analysis. Radiotherapy was not entered
because different radiotherapy regimes were used during
the 10-year period and because confounding was possi-
ble due to patient selection. The factor >4 PRC’s dur-
ing admission was not entered because patients who
received more than four PRC’s also suffered from a
post-operative haemorrhage. Therefore, only haemor-
rhage was entered. The AJCC/UICC 1997 classification
was not entered because this factor was represented by
the factor metastatic disease. Of these seven factors that
were entered in the multivariate analysis, only meta-
static disease (hazard ratio 1.43, CI 1.05–1.95) and
tumour size (hazard ratio 1.15, CI 1.01–1.31) indepen-
dently predicteda poor
summarised in Table 4 and Fig. 4.
The present study shows the outcome of a single-
centre analysis of the survival in 343 patientsafter surgical
treatment for histologically-proven pancreatic adeno-
carcinoma during a 10-year period between 1992 and
2002. Although resection is still the only treatment with
a curative potential, long-term survival was limited. The
median and 5-year survival after resection was only 17
months and 8%, and 7.5 months and zero after a sur-
gical bypass. Independent prognostic factors that pre-
dicted a poor survival for patients after resection were
tumour-positive lymph nodes, tumour-positive resection
margins, poor tumour differentiation grade and tumour
size. After bypass surgery, metastatic disease and
tumour size were independent prognostic factors.
The median survival after resection is comparable,
but 5-year survival is generally lower than rates reported
the literature (Table 5) [7–19]. The low 5-year survival
can partly be explained by the fact that in the present
study the follow-up of all patients was nearly com-
pleted. Only 22 (6.4%) of 343 patients were alive at the
end of the follow-up in April 2003 and only 1 patient
(0.3%) was lost to follow-up. Therefore, in this study
the actuarial survival computed by the Kaplan–Meier
method approaches the actual survival. When the
Kaplan–Meier method is used to perform survival cal-
culations, the percentage of 5-year survival increases
when increasing numbers of patients are censored or
lost to follow-up. This phenomenon occurs frequently in
series of patients with pancreatic carcinoma and was
described by Gudjonsson, who stated that survival
Patient numbers, study period, median survival, 5-year survival, number of 5-year survivors and mortality in large series of resections for pancreatic
carcinoma published in English literature since 1992
Study, journal, year [Ref.]Patient
No. of 5-year
Geer and colleagues, Am J Surg 1993 
Nitecki and colleagues, Ann Surg 1995 
Allema and colleagues, Cancer 1995 
Yeo and colleagues, Ann Surg 1995 
Sperti and colleagues, Br J Surg 1996 
Yeo and colleagues, Ann Surg 1997 
Millkan and colleagues, Am Surg 1999 
Wenger and colleagues, Dig Surg 2000 
Sohn and colleagues, J Gastroint Surg 2000 
van Geenen and colleagues, Eur J Surg Oncol 2001 
Kedra and colleagues, Hepatogastroenterology 2001 
Ahmad and colleagues, Am J Gastroenterol 2001 
Lim and colleagues, Ann Surg 2003 
This study1601992–2001 17.07.870
nst, not stated in article.
aHospital mortality (17 patients) excluded.
b3-year survival: 30%.
c3-year survival 34.3%.
K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558555
improves when the amount of censored data increases
. Yeo and colleagues, who reviewed 650 pancreati-
coduodenectomies in 1997, have endorsed this view-
point and state that their Kaplan–Meier curves lose
accuracy after a 2-year period because of a short period
of follow-up of 12 months . In most series published
in literature, the period of follow-up is not given and
information about censored patients is lacking. There-
fore, 5-year survival data should be interpreted with
some restraint. Certainly because the numbers of actual
5-year survivors, if stated, is mostly well below 10% of
the total number of studied patients (Table 5). When we
once again analyse the 67 patients treated in our centre
between 1983 and 1992 (published by Allema and col-
leagues in 1995 ), we find that one patient is lost-to-
follow up because he returned to his native country and
all other patients died. The 5-year survival of these 67
patients is 7.9%, while this was 15% in the original
article, indicating the effect of incomplete data. The
median and 5-year survival after bypass surgery is
comparable with literature, although few studies have
been published recently [5,25].
The median survival of patients after resection, which
was performed as a standard resection without extended
lymph node resection, is in agreement with the literature
during the last decade. Although resections were mac-
roscopically radical during surgery, our study shows a
high rate of positive resection margins after resection. In
these patients, the resection should be considered as a
palliative treatment. Concerning the margins of the
superior mesenteric artery or portal vein it is shown in
literature that a wedge resection does not lead to pro-
longed survival . The rate of tumour-positive resec-
tion margins at the pancreatic margin is high, certainly
because a re-resection of this margin can be performed
easily. Of the 31 patients with a tumour-positive pan-
creatic margin, in 12 patients this margin was the only
tumour-positive margin. Therefore, these patients could
theoretically have benefited from a frozen section and
subsequent pancreatic re-resection. As a result of these
findings, we have adapted our policy in terms of taking
frozen sections routinely.
In-hospital mortality after pancreaticoduodenectomy
is decreasing. Cameron  and Yeo  from Johns
Hopkins reported studies with series of 145 and 190
consecutive pancreaticoduodenectomies, without mor-
tality. However, these resections concerned hetero-
geneous groups of patients with malignant as well as
benign disease. To our best knowledge, the present
study is the first series with a zero hospital mortality in
160 consecutive resections for pancreatic adenocarci-
noma. In a previous study in our hospital of 176
patients who underwent resection between 1983 and
1992, mortality was 7.4% for patients who underwent a
resection for pancreatic carcinoma . Although there
was no mortality after resections for pancreatic adeno-
carcinoma, the overall mortality for the total group of
patients after pancreaticoduodenectomy was 1.5% in
the period from 1992 to 1998 . Low mortality was
found after resections for pancreatitis and pancreatic
carcinoma and a higher mortality after resections for
ampullary and bile duct carcinoma. This difference
could be due to the lower incidence of major compli-
cations like anastomotic leakage, in which the quality of
the pancreatic remnant and dilatation of the pancreatic
duct plays an important role . Furthermore, the
preoperative selection procedure may play a role. The
selection of patients with tumours other than pancreatic
carcinomas could have been less stringent, because of
their better prognosis. The mortality after palliative
bypass was low (2%).
Morbidity after pancreaticoduodenectomy is also
decreasing, our rates for the three major surgery-related
complications after a pancreaticoduodenectomy were
for anastomotic leakage 7%, haemorrhage 6% and
abdominal abscess 10%. This is in agreement with the
Irresectability of the tumour occurred in 53% of the
patients and was mostly assessed during explorative
laparotomy. Although less invasive palliative treatments
such as stenting and laparoscopic bypass procedures are
possible and are in fact first choice when irresectability
has been shown at radiological staging, bypass surgery
at the time of laparotomy is efficient to prevent later
biliary and duodenal obstruction . A recent rando-
mised study which compared bypass surgery with
endoscoping stenting showed no benefit for patients
allocated to the endoscopic treatment. Furthermore,
survival after bypass surgery was longer while proce-
dure-related morbidity and mortality were comparable
. In the present study, morbidity after bypass
surgery is low and hospital stay is relatively short. In
combination with an acceptable hospital mortality, a
surgical bypass is a safe and efficient treatment option
for patients with irresectable pancreatic adenocarci-
noma found at explorative laparotomy.
In the past, the role of an additional prophylactic
gastrojejunostomy has been discussed. Lillemoe and
colleagues demonstrated in a randomised trial that a
gastrojejunostomy supplementary to biliary bypass
decreases the incidence of late gastric outlet obstruction,
without a higher mortality and morbidity . A rando-
mised trial in our centre between 1999 and 2002 shows
comparable results . In the present study, most
patients (84%) with an irresectable tumour underwent a
double bypass. Thirteen patients (7.1%) needed re-
operation after bypass surgery because of late gastro-
intestinal obstruction (7/157 after double bypass and
6/26 after single bypass).
The independent prognostic factors found in the
resection group were related to tumour characteristics.
During the last decade, various studies showed tumour-
556K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558
related prognostic factors after resection [7–19]. There-
fore, the biology of pancreatic adenocarcinoma seems
responsible for the poor prognosis. In the light of the
finding that tumour-positive resection margins and
lymph nodes are independent prognostic factors, one
could consider to study the effect of an extended lymph
node resection in a large randomised controlled trial.
However, two single centre randomised trials could not
show a survival benefit after an extended lymph node
resection [32,33]. Chemo-radiation therapy has also
been found to be an independent factor for survival in
the literature [11;12;19]. However, patient selection may
be responsible for this finding in these retrospective
studies. In the bypass group, only metastatic disease and
tumour size were independently prognostic for survival.
These factors are also related to the aggressiveness of
the tumour and the stage of the disease.
Although mortality and morbidity have decreased
rapidly after pancreatic surgery, survival of pancreatic
cancer remains poor. Therefore, new treatment mod-
alities, such as immunotherapy, molecularly-targeted
therapies and gene therapy, are crucial to specifically
target pancreatic adenocarcinoma. These new therapies
must reach the primary tumour and, even more impor-
tantly, must reach distant metastases and micro-
metastases. A combined eradication of both the primary
tumour and the (micro)metastatic tissue is the only way
to improve the survival of these patients with a poor
All authors certify that there are no affiliations with
or involvement in any organisation, person or entity
with a direct financial interest in the subject matter or
materials discussed in this manuscript.
1. Greenlee RT, Murray T, Bolden S, et al. Cancer statistics, 2000.
CA Cancer J Clin 2000, 50, 7–33.
2. Sener SF, Fremgen A, Menck HR, et al. Pancreatic cancer: a
report of treatment and survival trends for 100,313 patients
diagnosed from 1985–1995, using the National Cancer Database.
J Am Coll Surg 1999, 189, 1–7.
3. Tan HP, Smith J, Garberoglio CA. Pancreatic adenocarcinoma:
an update. J Am Coll Surg 1996, 183, 164–184.
4. Smith AC, Dowsett JF, Russell RC, et al. Randomised rial of
endoscopic stenting versus surgical bypass in malignant low bile-
duct obstruction. Lancet 1994, 344, 1655–1660.
5. Lillemoe KD, Cameron JL, Hardacre JM, et al. Is prophylactic
gastrojejunostomy indicated for unresectable periampullary can-
cer? A prospective randomized trial. Ann Surg 1999, 230, 322–328.
6. Lillemoe KD, Cameron JL, Kaufman HS, et al. Chemical
splanchnicectomy in patients with unresectable pancreatic cancer.
A prospective randomized trial. Ann Surg 1993, 217, 447–455.
7. Geer RJ, Brennan MF. Prognostic indicators for survival after
resection of pancreatic adenocarcinoma. Am J Surg 1993, 165,
8. Nitecki SS, Sarr MG, Colby TV, et al. Long-term survival after
resection for ductal adenocarcinoma of the pancreas. Is it really
improving? Ann Surg 1995, 221, 59–66.
9. Sperti C, Pasquali C, Piccoli A, et al. Survival after resection for
ductal adenocarcinoma of the pancreas. Br J Surg 1996, 83, 625–
10. Wenger FA, Peter F, Zieren J, et al. Prognosis factors in carci-
noma of the head of the pancreas. Dig Surg 2000, 17, 29–35.
11. Ahmad NA, Lewis JD, Ginsberg GG, et al. Long term survival
after pancreatic resection for pancreatic adenocarcinoma. Am J
Gastroenterol 2001, 26, 2609–2615.
12. Lim JE, Chien MW, Earle CC. Prognostic factors following
curative resection for pancreatic adenocarcinoma: a population-
based, linked database analysis of 396 patients. Ann Surg 2003,
13. van Geenen RC, van Gulik TM, Offerhaus GJ, et al. Survival
after pancreaticoduodenectomy for periampullary adenocarci-
noma: an update. Eur J Surg Oncol 2001, 27, 549–557.
14. Yeo CJ, Cameron JL, Sohn TA, et al. Six hundred fifty con-
secutive pancreaticoduodenectomies in the 1990s: pathology,
complications, and outcomes. Ann Surg 1997, 226, 248–257.
15. Allema JH, Reinders ME, van Gulik TM, et al. Prognostic fac-
tors for survival after pancreaticoduodenectomy for patients with
carcinoma of the pancreatic head region. Cancer 1995, 75, 2069–
16. Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduode-
nectomy for cancer of the head of the pancreas. 201 patients. Ann
Surg 1995, 221, 721–731.
17. Kedra B, Popiela T, Sierzega M, et al. Prognostic factors of long-
term survival after resective procedures for pancreatic cancer.
Hepatogastroenterology 2001, 48, 1762–1766.
18. Millikan KW, Deziel DJ, Silverstein JC, et al. Prognostic factors
associated with resectable adenocarcinoma of the head of the
pancreas. Am Surg 1999, 65, 618–623.
19. Sohn TA, Yeo CJ, Cameron JL, et al. Resected adenocarcinoma
of the pancreas-616 patients: results, outcomes, and prognostic
indicators. J Gastrointest Surg 2000, 4, 567–579.
20. Obertop H, Gouma DJ. Improved results for resection of peri-
ampullary adenocarcinoma. HPB Surg 1996, 10, 125–128.
21. van Geenen RC, ten Kate FJ, de Wit LT, et al. Segmental resec-
tion and wedge excision of the portal or superior mesenteric vein
during pancreatoduodenectomy. Surgery 2001, 129, 158–163.
22. Gouma DJ, van Gulik TM, de Wit LT, et al. Complications after
resection of biliopancreatic cancer. Ann Oncol 1999, 10(Supp. 4),
23. Obertop H, Pedrazzoli S. Current views on surgical treatment of
pancreatic cancer. Dig Surg 1999, 16, 263–264.
24. Gudjonsson B. Survival statistics gone awry: pancreatic cancer, a
case in point. J Clin Gastroenterol 2002, 35, 180–184.
25. Engelken FJ, Bettschart V, Rahman MQ, et al. Prognostic fac-
tors in the palliation of pancreatic cancer. Eur J Surg Oncol 2003,
26. Cameron JL, Pitt HA, Yeo CJ, et al. One hundred and forty-five
consecutive pancreaticoduodenectomies without mortality. Ann
Surg 1993, 217, 430–435.
27. Yeo CJ, Cameron JL, Lillemoe KD, et al. Does prophylactic
octreotide decrease the rates of pancreatic fistula and other
complications after pancreaticoduodenectomy? Results of a
prospective randomized placebo-controlled trial. Ann Surg 2000,
28. Halloran CM, Ghaneh P, Bosonnet L, et al. Complications of
pancreatic cancer resection. Dig Surg 2002, 19, 138–146.
29. van den Bosch RP, van der Schelling GP, Klinkenbijl JHMPG, et
al. Guidelines for the application of surgery and endoprostheses
K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558557
in the palliation of obstructive jaundice in advanced cancer of the
pancreas. Ann Surg 1994, 219, 18–24.
30. Nieveen van Dijkum EJ, Romijn MG, Terwee CB, et al.
Laparoscopic staging and subsequent palliation in patients with
peripancreatic carcinoma. Ann Surg 2003, 237, 66–73.
31. van Heek NT, de Castro SMM, van Eijck CH, et al. The need for
a prophylactic gastrojejunostomy for unresectable periampullary
cancer, a prospective randomized multi-center trial with special
focus on assessment of quality of life. Ann Surg 2003, 238, 894–
32. Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduode-
nectomy with or without distal gastrectomy and extended retro-
peritoneal lymphadenectomy for periampullary adenocarcinoma,
part 2: randomized controlled trial evaluating survival, morbid-
ity, and mortality. Ann Surg 2002, 236, 355–366.
33. Pedrazzoli S, DiCarlo V, Dionigi R, et al. Standard versus
extended lymphadenectomy associated with pancreatoduode-
nectomy in the surgical treatment of adenocarcinoma of the head
of the pancreas: a multicenter, prospective, randomized study.
Lymphadenectomy Study Group. Ann Surg 1998, 228, 508–517.
558 K.F.D. Kuhlmann et al./European Journal of Cancer 40 (2004) 549–558